The Onset of the Volcanism in the Ciomadul Volcanic Dome Complex (Eastern Carpathians): Eruption Chronology and Magma Type Variation

Home , Ciomadul, Shoshonite, Slab (geology)

Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Contents lists available at ScienceDirect

Journal of Volcanology and Geothermal Research

journal homepage: www.elsevier.com/locate/jvolgeores

The onset of the volcanism in the Ciomadul Volcanic Dome Complex (Eastern Carpathians): Eruption chronology and magma type variation

Kata Molnár a,⁎, Szabolcs Harangi a,b, Réka Lukács b,IstvánDunklc, Axel K. Schmitt d, Balázs Kiss b, Tamás Garamhegyi e,IoanSeghedif a Department of Petrology and Geochemistry, Eötvös Loránd University, Pázmány Péter stny. 1/c, H-1117 Budapest, Hungary b MTA-ELTE Volcanology Research Group, Budapest, Hungary c Sedimentology and Environmental Geology, Geoscience Centre, Georg-August University, Göttingen, Germany d Institute of Earth Sciences, Ruprecht-Karls University, Heidelberg, Germany e Department of Physical and Applied Geology, Eötvös Loránd University, Budapest, Hungary f Institute of Geodynamics, Romanian Academy, Bucharest, Romania article info abstract

Article history: Combined zircon U-Th-Pb and (U-Th)/He dating was applied to reﬁne the eruption chronology of the last 2 Myr Received 29 June 2017 for the andesitic and dacitic Pilişca volcano and Ciomadul Volcanic Dome Complex (CVDC), the youngest volcanic Received in revised form 9 January 2018 area of the Carpathian-Pannonian region, located in the southernmost Harghita, eastern-central Europe. The pro- Accepted 29 January 2018 posed eruption ages, which are supported also by the youngest zircon crystallization ages, are much younger Available online 3 February 2018 than the previously determined K/Ar ages. By dating every known eruption center in the CVDC, repose times be-

Keywords: tween eruptive events were also accurately determined. Eruption of the andesite at Murgul Mare (1865 ± 87 ka) ş Zircon (U-Th)/He dating and dacite of the Pili ca volcanic complex (1640 ± 37 ka) terminated an earlier pulse of volcanic activity within Eruption chronology the southernmost Harghita region, west of the Olt valley. This was followed by the onset of the volcanism in the U-series dating CVDC, which occurred after several 100s kyr of eruptive quiescence. At ca. 1 Ma a signiﬁcant change in the com- Quaternary position of erupted magma occurred from medium-K calc-alkaline compositions to high-K dacitic (Baba-Laposa Harghita dome at 942 ± 65 ka) and shoshonitic magmas (Malnaş and Bixad domes; 964 ± 46 ka and 907 ± 66 ka, respec- Carpathian-Pannonian region tively). Noteworthy, eruptions of magmas with distinct chemical compositions occurred within a restricted area, a few km from one another. These oldest lava domes of the CVDC form a NNE-SSW striking tectonic lineament along the Olt valley. Following a brief (ca. 100 kyr) hiatus, extrusion of high-K andesitic magma continued at Dealul Mare (842 ± 53 ka). After another ca. 200 kyr period of quiescence two high-K dacitic lava domes extruded (Puturosul: 642 ± 44 ka and Balvanyos: 583 ± 30 ka). The Turnul Apor lava extrusion occurred after a ca. 200 kyr repose time (at 344 ± 33 ka), whereas formation of the Haramul Mic lava dome (154 ± 16 ka) represents the onset of the development of the prominent Ciomadul volcano. The accurate determination of eruption dates shows that the volcanic eruptions were often separated by prolonged (ca. 100 to 200 kyr) quiescence periods. Demonstration of recurrence of volcanism even after such long dormancy has to be considered in assessing volcanic hazards, particularly in seemingly inactive volcanic areas, where no Holocene eruptions occurred. The term of ‘volcanoes with Potentially Active Magma Storage’ illustrates the potential of volcanic rejuvenation for such long-dormant volcanoes with the existence of melt-bearing crustal magma body. © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction eruptions N10 kyr and possibly long repose times between active phases. Therefore, it is not easy to evaluate whether these volcanoes Constraining eruption chronology has a primary importance in can be reactivated after prolonged quiescence or have become extinct assessing volcanic hazards, since it provides information about the fre- (Connor et al., 2006). The poor knowledge about the nature of such quency of eruptions and the length of the repose time. Databases such long-dormant volcanoes is partly due to the difﬁculties to apply appro- as the Smithsonian's Global Volcanism Program collect available data priate geochronometers, which can be used to accurately determine the primarily for the Holocene volcanic activity (Siebert et al., 2011). On eruption ages for the last 1 Myr. the contrary, there is much less knowledge on volcanoes with known Over the last decade, combined application of U-Pb or U-Th and (U- Th)/He zircon geochronology has become a promising method to date ⁎ Corresponding author. volcanic eruptions occurring during the second half of Pleistocene E-mail address: [email protected] (K. Molnár). epoch (e.g., Schmitt et al., 2006, 2010; Danišík et al., 2012, 2017;

https://doi.org/10.1016/j.jvolgeores.2018.01.025 0377-0273/© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 40 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Harangi et al., 2015a; Mucek et al., 2017). This technique is proved to be allows for an independent check of the accuracy of the dating. The erup- particularly applicable when other dating methods such as radiocarbon, tion ages were combined by newly analyzed major and trace element K/Ar or 40Ar/39Ar techniques encounter analytical or interpretational compositional data allowing the evaluation of chemical changes in the difficulties often caused by a lack of appropriate materials for dating erupted magmas. These data yield important new information about (Danišík et al., 2017). Rapid temperature decrease below the ~200 °C the nature of a little known long-dormant volcano in eastern-central closure temperature for He diffusion in zircon enables determination Europe. of eruption ages based on the (U-Th)/He systematics, assuming that no subsequent heating occurred (Farley, 2002). 2. Geological background The youngest volcanic activity of the Carpathian-Pannonian region occurred at Ciomadul volcano (Eastern Carpathians, Romania; Fig. 1). The Călimani-Gurghiu-Harghita volcanic chain (CGH) is located in Although, its youngest eruption is dated at ca. 30 ka (Vinkler et al., the southeastern part of the Carpathian-Pannonian region (CPR), 2007; Harangi et al., 2010, 2015a; Karátson et al., 2016) the composition eastern-central Europe (Fig. 1). The CPR comprises the Pannonian and flux of the emitted gases at mofettas (Vaselli et al., 2002; Kis et al., basin characterized by thin continental crust and lithosphere 2017), seismic tomography (Popa et al., 2012)aswellascombinedpet- surrounded by orogenic mountain belts of the Alps, Carpathians and rologic and magnetotelluric studies (Kiss et al., 2014; Harangi et al., Dinarides (e.g., Horváth et al., 2006). Major thinning of the lithosphere 2015b) indicate that there is still melt-bearing magma body beneath has been considered to be a result of the Early to Mid-Miocene the volcanic complex and therefore, there is still a potential for further retreating subduction of an oceanic lithosphere underneath at the East- volcanic activity at Ciomadul (Szakács et al., 2002; Harangi, 2007; ern Carpathians (e.g., Royden et al., 1983; Csontos et al., 1992; Horváth Szakács and Seghedi, 2013). Thus, there is a requirement for an accurate et al., 2006). The southeastern edge of the CGH is located ~50 km from knowledge on the eruption behavior of the volcanic system. The timing the Vrancea zone that exhibits the largest present-day strain concentra- of the last explosive eruptions and the chronology of the latest volcanic tion in continental Europe (Wenzel et al., 1999; Ismail-Zadeh et al., phase was determined by several authors (i.e., Moriya et al., 1995, 1996; 2012). This is due to a near-vertical slab descending into the upper man- Vinkler et al., 2007; Harangi et al., 2010, 2015a; Karátson et al., 2016). In tle (Oncescu et al., 1984; Martin et al., 2006) that poses a persistent re- this paper, we focus on the older history, i.e., the onset of the volcanism gional seismic hazard. The reason of the descending vertical lithospheric at the Ciomadul Volcanic Dome Complex (cf. Szakács et al., 2015), which slab beneath Vrancea is strongly debated. Explanations involve as it rep- comprises the volcanic edifice of Ciomadul and the peripheral lava resents the latest stage of the subduction along the Carpathians domes. Based on previous K/Ar ages (Peltz et al., 1987; Pécskay et al., (Sperner et al., 2001), but lithospheric delamination in the absence of 1995; Szakács et al., 2015) the peripheral domes are considered to subduction (Fillerup et al., 2010; Koulakov et al., 2010) or a combination have formed between 2 and 0.5 Ma. Here, we adopted the methodology of lithospheric delamination and subduction roll-back have been alter- used by Schmitt et al. (2006), Danišík et al. (2012) and Harangi et al. natively proposed (Gîrbacea and Frisch, 1998; Girbacea et al., 1998; (2015a), where they determined eruption ages from combined U-Th Chalot-Prat and Gîrbacea, 2000). and (U-Th)/He zircon dating. Individually, these chronometers yield in- Volcanic activity of the South Harghita and the nearby Perşani alkali dependent constraints on crystallization and eruption ages, but because basalt volcanic field could be related to this active tectonic setting, al- zircon crystallization always precedes eruption, the combined methods though the link between the vertical slab movement, magma

Fig. 1. (A) Simplified geological map about the location of the Călimani-Gurghiu-Harghita volcanic chain in the Carpathian-Pannonian region (modified after Cloetingh et al., 2004 and Harangi et al., 2013). The rectangle indicates the area of the South Harghita. (B) Geological map of the South Harghita with the sampling sites (based on the 1:200,000 map of Ianovici and Rădulescu, 1966, Geological Institute of Romania and Szakács et al., 2015). G: Gurghiu Mts., H: Harghita Mts.; PD: Pilişca, BL: Baba-Laposa, AB: Turnul Apor, KHM: Haramul Mic, BH: Puturosul, BAL: Balvanyos, NH: Dealul Mare, MB-M: Malnaş, MB-B: Bixad, NM: Murgul Mare. K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 41 generation, and volcanism is still unclear (Downes et al., 1995; Harangi dacitic Ciomadul Volcanic Dome Complex (Casta, 1980; Peltz et al., et al., 2013; Seghedi et al., 2004, 2011, 2016). 1987; Seghedi et al., 1987; Szakács et al., 1993, 2015; Pécskay et al., In the CPR, eruptions of various magmas from basalt to rhyolite have 1995, 2006; Fig. 1B; Table 1). Chemical composition of the chain end- occurred since 18 Ma (Szabó et al., 1992; Harangi, 2001; Konečný et al., member Ciomadul volcanics was interpreted to mark the transition 2002; Seghedi et al., 2004, 2011; Harangi and Lenkey, 2007; Seghedi from normal calc-alkaline to adakite-like magmas (Seghedi et al., 2011). and Downes, 2011). In the eastern part, volcanism occurred at the The Ciomadul Volcanic Dome Complex comprises the massive Călimani-Gurghiu-Harghita andesitic-dacitic volcanic chain from ca. Ciomadul volcano and the surrounding peripheral lava domes (Dealul 10.2 Ma onward and gradually shifted to the southeast with waning in- Mare, Haramul Mic, Puturosul and Balvanyos; Fig. 1B; Table 1) as de- tensity (Peltz et al., 1987; Pécskay et al., 1995, 2006; Szakács et al., fined by Szakács et al. (2015). However, we consider that the two 1997). The CGH volcanism postdates the inferred subduction during shoshonitic domes at Malnaş and Bixad as well as the dacitic Baba- the mid-Miocene (Cloetingh et al., 2004) and therefore is considered Laposa dome also belong to this volcanic system. The volcanic edifices post-collisional (Mason et al., 1996; Seghedi et al., 1998; Seghedi and were constructed on the Cretaceous Ceahlău-Severin flysch nappe and Downes, 2011). Volcanic activity in the South Harghita (Luci-Lazu, are bordered by the Plio-Pleistocene intramontane Lower Ciuc basin in Cucu, Pilişca and Ciomadul; Fig. 1) started at 5.3 Ma (Pécskay et al., the north (Fig. 1B). The youngest phase of volcanism occurred at the 1995) and was coeval with or slightly post-dated the subsidence of Ciomadul volcano between ~56 and 32 ka based on radiocarbon, ther- the nearby Braşov, Gheorgheni and Ciuc basins (Girbacea et al., 1998; mal luminescence and combined zircon U-Th and (U-Th)/He dating Fielitz and Seghedi, 2005). After a gap between 3.9 and 2.8 Ma (e.g., Moriya et al., 1995, 1996; Vinkler et al., 2007; Harangi et al., (Pécskay et al., 2006; Seghedi et al., 2011), a sharp compositional 2010, 2015a; Karátson et al., 2016). It represents the youngest volcanic change occurred in the erupted magmas, which is reflected in the activity of the entire Carpathian-Pannonian region. more potassic and incompatible element-enriched nature at Cucu, Pilişca and Ciomadul compared to earlier volcanism (Mason et al., 3. Samples and analytical methods 1996; Harangi and Lenkey, 2007; Seghedi et al., 2011). This transition spatially coincides with the southward migration of the volcanic centers 3.1. Sampling and whole-rock techniques across the Trotuş fault (Fig. 1). The Trotus fault tectonically separates the European and Moesian continental blocks, which corresponds to a This study focuses mainly on the peripheral lava domes (Szakács decrease in crustal thickness from 40–45 km in the European to et al., 2015) surrounding the central lava dome edifice of Ciomadul 35–40 km in the Moesian block (e.g., Cloetingh et al., 2004). (Fig. 1B). They represent the older phase of the volcanism in the The southern, more potassic volcanism of South Harghita developed CVDC. We attempted to sample all the known eruptive centers in addi- between 2.8 and 0.03 Ma and comprises the andesitic Cucu, the basaltic- tion to sampling the youngest volcanic products of the nearby Pilişca andesitic, andesitic and dacitic Pilişca volcanoes and the andesitic and volcano. Sample sites, names and lithology are listed in Table 1;the

Table 1 Sample names, localities, lithologies and previous age results of the studied samples from the southernmost Harghita. The corresponding Hungarian names of the locations are also indicated in italic.

Sample code in Sample code (in Location GPS Lithology “Afﬁliation” and Dated Age (Ma) Method Reference the text the lab) coordinates reference fraction

NM CSO-NM1 Murgul Mare - Nagy 46°4′ 25°48′ Andesite Pilişca - Panaiotu Whole 2.69 ± K/Ar Szakács et al. Murgó 36″ 01″ et al. (2012) rock 0.22 (2015) Whole 2.25 ± rock 0.09 PD CSO-PD2 Pilişca - Piliske 46° 9′ 25°50′ Dacite Pilişca - Seghedi et al. Biotite 2.1 ± 0.1 K/Ar Pécskay et al. 16″ 47″ (1987) (1995) Whole 1.8 ± 0.1 K/Ar Pécskay et al. rock (1995) Biotite 1.5 ± 0.1 K/Ar Pécskay et al. (1995) MB-M CSO-MB-M Malnaş quarry - 46°2′ 25°48′ Shoshonite Distinct unit - Seghedi Whole 2.22 ± K/Ar Peltz et al. Málnás 59″ 43″ et al. (1987) rock 0.14 (1987) 1.95–1.77 Palaeomagnetic Panaiotu et al. (2012) MB-B CSO-MB-B Bixad quarry - 46°5′ 25°50′ Shoshonite Distinct unit - Seghedi Whole 1.45 ± K/Ar Peltz et al. Bükszád 1″ 1″ et al. (1987) rock 0.40 (1987) 1.95–1.77 Palaeomagnetic Panaiotu et al. (2012) BL CSO-BL Baba-Lapoşa- 46°10′ 25°52′ Dacite Pilişca - Szakács et al. Whole 1.46 ± K/Ar Szakács et al. Bába-Laposa 33″ 27″ (2015) rock 0.27 (2015) NH CSO-NH5 Dealul Mare - 46°6′ 25°55′ Andesite Ciomadul - Panaiotu Whole 1.02 ± K/Ar Szakács et al. Nagy-Hegyes 06″ 9″ et al. (2012) rock 0.07 (2015) BH CSO-BH Puturosul - 46°7′ 25°56′ Dacite Ciomadul - Panaiotu Whole 0.71 ± K/Ar Szakács et al. Büdös-hegy 11″ 55″ et al. (2012) rock 0.04 (2015) CSO-BH(sz) 46°7′ 25°56′ 8″ 56″ BAL MK-2 Cetatea Balvanyos - 46°6′ 25°57′ Dacite Ciomadul - Panaiotu Whole 1.02 ± K/Ar Pécskay et al. Bálványos 42″ 53″ et al. (2012) rock 0.15 (1995) CSO-BAL(cs) 46°6′ 25°58′ Whole 0.92 ± K/Ar Pécskay et al. 54″ 4″ rock 0.18 (1995) AB CSO-AB Turnul Apor - 46°8′ 25°51′ Dacite ––––– Apor-bástya 6″ 51″ KHM CSO-KHM1 Haramul Mic - 46°10′ 25°55′ Dacite Ciomadul - Seghedi Whole 0.85 K/Ar Casta (1980) Kis-Haram 38″ 25″ et al. (1987) rock 42 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Hungarian names of each locality are also indicated due to the bilingual In order to perform initial Th disequilibrium correction for the (U- character of this part of Romania. Th)/He ages, zircon crystals from all samples were dated by in-situ U- Petrology of the studied samples and textural characterization of zir- Th-Pb geochronology using LA-ICP-MS (ETH Zürich) and/or SIMS (HIP con crystals were performed by combined investigation with petro- laboratory, University of Heidelberg) methods. Analytical backgrounds graphic microscope and an AMRAY 1830 type SEM + EDAX PV9800 and results are presented in Lukács et al. (submitted).Incaseofthe EDS equipped with a GATAN MiniCL at the Department of Petrology KHM sample (the youngest sample dated in this study), double-dating and Geochemistry of the Eötvös Loránd University, Budapest, Hungary. was also conducted, i.e., after in-situ SIMS analysis, 4 zircon crystals Whole-rock major and trace element geochemical compositions from the indium mount were picked out, cleaned by cc. HCl and ana- were analyzed at AcmeLabs (Vancouver, Canada; http://acmelab.com/ lyzed for (U-Th)/He ages. ). Major and minor elements were determined by ICP-emission spec- trometry and trace elements were analyzed by ICP-MS following a lith- 4. Results ium borate fusion and dilution in acid. 4.1. General petrography and geochemistry 3.2. Zircon extraction

The studied lava dome rocks comprise andesites (SiO2 =58–62 wt%) Approximately 1 kg of rock was collected from each outcrop. Sam- and dacites (SiO2 =63–67 wt%; Fig. 2A) with calc-alkaline, high-K calc- ples were crushed and sieved, and the 63–125 μm fraction was gravity alkaline and shoshonitic composition (Table 2). Three groups can be dis- separated in heavy liquid (sodium polytungstate with a density of tinguished based on their SiO2 and K2Ocontents(Fig. 2A) according to the 2.88 ± 0.01 g/cm3). Zircon was concentrated by removing Fe-Ti oxides classiﬁcation of Peccerillo and Taylor (1976): using a permanent magnet. Inclusion- and ﬁssure-free intact, euhedral – ş zircon crystals N60 μm in width were hand-picked from the non- (1) medium-K calc-alkaline series (CAs) Murgul Mare and Pili ca – magnetic fraction under a binocular microscope, photographed and (2) high-K calc-alkaline series (HKCAs) Baba-Laposa, Haramul Mic, packed into platinum capsules for He degassing. For the petrographic Puturosul, Balvanyos, Dealul Mare and Turnul Apor – ş and textural characterization of zircon populations, ~80 crystals per (3) high-K calc-alkaline-shoshonitic series (HKSs) Malna and sample were hand-picked, mounted in a 2.54 cm epoxy disk and Bixad. polished.

3.3. (U-Th)/He-analysis The volcanic rocks are crystal-rich, porphyritic (phenocryst content is 25–40 vol%) with abundant glomerocrystic aggregates and crystal (U-Th)/He age determinations were carried out at the GÖochron clots in a hyalopilitic to holocrystalline matrix. The phenocryst assem- Laboratory of Georg-August University, Göttingen. Single-crystal ali- blage of the CAs and HKCAs samples invariably comprises plagioclase, quots were dated, usually six aliquots per sample. The Pt capsules amphibole and biotite. Additionally, clinopyroxene and rounded quartz were heated for 5 min (ca. 1200 °C) under high vacuum by an infra- are present in the Murgul Mare andesite (CAs), and clinopyroxene, red laser and the extracted gas was purified using a SAES Ti-Zr getter orthopyroxene and olivine in the Turnul Apor dacite (HKCAs). The dom- at 450 °C. The chemically inert noble gases and a minor amount of inantly high-Mg character of these minerals and their occurrence in other rest gases were then expanded into a Hiden triple-filter quadru- microenclaves, crystal clots, glomerocrysts and individual crystals are pole mass spectrometer equipped with a positive ion counting detector. similar to the younger lava dome rocks of the Ciomadul volcano. The ac- One or more “gas re-extract” steps were run for each sample to verify cessory phases are zircon and apatite, while the HKCAs samples contain complete degassing of the crystals. The residual gas is usually around additional titanite. Most of the samples have a fine-grained groundmass

1 to 2% after the ﬁrst extraction. He gas results were corrected for with plagioclase microlites, Fe-Ti oxides and SiO2 batchesexceptforthe blank, determined by heating empty Pt capsules using the same Murgul Mare andesite which has a coarse-grained groundmass. procedure. The samples of the HKSs contain much fewer phenocrysts (5–10%) Following degassing, zircon crystals were retrieved from the gas ex- which comprise dominantly clinopyroxene and altered feldspar, whereas traction line, extracted from the Pt capsules and spiked with calibrated biotite, amphibole together with rounded quartz, orthopyroxene, olivine 230Th and 233U solutions. The crystals were dissolved in pressurized and titanite are present in minor amounts. In these rocks the groundmass Savillex teﬂon bombs using a mixture of double distilled 48% HF and is coarse-grained, containing dominantly feldspar and clinopyroxene

65% HNO3 at 220 °C during 5 days. Spiked solutions were analyzed as microlites, the accessories are zircon, apatite and titanite. The mineral 0.4 ml solutions by isotope dilution for 233U, 235U, 238U, 230Th, 232Th phases of both the HKCAs and HKSs samples together with the Murgul and 147Sm using a Perkin Elmer Elan DRC II ICP-MS with an APEX Mare sample (CAs) display diverse, disequilibrium textures similar to micro-flow nebulizer. Procedural U and Th blanks by this method are the HKCAs dacite of the Ciomadul volcano implying open system usually very stable in a measurement session and below 1.5 pg. Total an- magma chamber processes involving mixing of distinct magmas (Kiss alytical uncertainty (TAU) was computed as the square root of the sum et al., 2014). of the squares of weighted uncertainties of U, Th, Sm and He measure- Earlier bulk rock geochemical data for the South Harghita volcanic ments. The raw (U-Th)/He ages were adjusted according to the alpha formations were reported by Szakács and Seghedi (1986), Mason et al. ejection correction procedure (FT) assuming homogeneous distribution (1996, 1998), Harangi and Lenkey (2007), Vinkler et al. (2007) and of U and Th and using the base equations of Farley et al. (1996) and Seghedietal.(2011). These were completed with new bulk rock com- Farley (2002) and the tetragonal crystal model and fit parameters positional data (Appendix 1). Most of the volcanic rocks (HKSs and given by Hourigan et al. (2005). The individual (U-Th)/He ages were cal- HKCAs) show enrichment of LIL (large ion lithophile; e.g., Cs, Rb, Ba, K, culated by the Taylor Expansion Method (Braun et al., 2012). The TAU Sr) and depletion of HFS (high field strength; e.g., Nb, Ti) elements and the estimated error of FT were used to calculate the uncertainties that are characteristic also of the youngest pumices of the Ciomadul vol- of FT-corrected (U-Th)/He ages. cano (Vinkler et al., 2007; Fig. 2C, D). They are particularly enriched in Sr The accuracy of zircon (U-Th)/He dating was monitored by replicate and Ba (N1000 ppm), whereas another common feature is the absence analyses of the reference material of Fish Canyon Tuff zircon, yielding a of negative Eu anomaly (Eu/Eu* values ranging from 1.1 to 0.9). The mean (U-Th)/He age of 28.1 ± 2.3 Ma (n = 128; where n is the number three compositional groups are clearly separated in the Sr vs Sm/Yb of replicate analyses per sample) which consistent with the reference plot (Fig. 2B). Furthermore, this plot illustrates also the distinct charac- (U-Th)/He age of 28.3 ± 1.3 Ma (Reiners, 2005). ters of the HKSs samples (Fig. 2B), possibly implying different K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 43

Fig. 2. Major and trace element variation diagrams (A and B) and trace element characteristics (C and D) of the studied samples. (A) SiO2 vs K2Oclassification diagram (the fields after Peccerillo and Taylor, 1976) for the studied rocks. The different grey fields indicate whole rock data from the Harghita for comparison from Mason et al. (1996) and Harangi and Lenkey (2007); (B) Sr vs Sm/Yb; (C) primitive mantle-normalized trace element plots and (D) chondrite-normalized REE plots (normalizing values after Sun and McDonough, 1989 and Nakamura, 1974, respectively). proportions of the mixed magmas (Seghedi et al., 1987; Mason et al., grains shown in Fig. 3. In the CAs samples, and in most of the HKCAs 1996). The geochemical difference between the samples of Murgul samples euhedral crystals are dominating, whereas in the HKSs- Mare, Malnaş and Bixad is remarkable, since they occur in the same re- samples and the HKCAs dacite of Puturosul and Balvanyos, zircon crys- stricted area, very close to each other (Fig. 1B). The two CAs samples, i.e. tals are usually rounded. The dimensions of the studied zircons vary the Murgul Mare and Pilişca are similar in Sr and Sm/Yb, but are distinct considerably (Table 3). Zircon crystals in samples Pilişca and Murgul in other major and trace element character. The Murgul Mare andesite Mare are the longest; they range from 320 to 780 μm and 130 to 620 shows some geochemical similarities to the HKCAs, whereas the Pilişca μm, respectively, whereas the smallest crystals are in the Dealul Mare dacite is clearly distinct and resembles the intermediate rocks of South andesite, where the zircon size varies between 120 and 200 μm. In Harghita (Mason et al., 1996). other samples, zircon crystals have a maximum length of 280 μm and a minimum length of 140 μm. 4.2. Characteristics of zircon crystals The average aspect ratio (length/width) is ~3:1, the most elongated crystals are in the samples of Pilişca and Balvanyos with an aspect ratio All of the studied samples contain zircon as accessory phase. The of ~5:1. Although the shape, size and appearance of the zircon crystals crystals occur as inclusions in the phenocrysts and in fewer amounts vary through the sample sets, their internal structure is uniform. They in the groundmass. All samples comprise dominantly euhedral crystals, typically show oscillatory zoning in each sample which is sometimes ac- but partially resorbed crystals are also present. Selected types of zircon companied by sector zoning (Fig. 3D).

Table 2 General petrological and geochemical features of the southernmost Harghita samples.

Arc rock-type Lithology Sample Phenocryst assemblage Accessories SiO2 (wt%) K2O (wt%) Normal calc-alkaline Andesite NM Plagioclase, amphibole, biotite, clinopyroxene, quartz Zircon, apatite 61.8 2.4 Dacite PD Plagioclase, amphibole, biotite Zircon, apatite 66.6 2.7 High-K calc-alkaline Andesite NH Amphibole, plagioclase, biotite, apatite Zircon, apatite, titanite 58.9 2.6 Dacite BL Plagioclase, biotite, amphibole, titanite Zircon, apatite, titanite 63.2–65.8 3.2–3.3 BH Plagioclase, amphibole, biotite, apatite Zircon, apatite, titanite BAL Plagioclase, biotite, amphibole, Zircon, apatite, titanite AB Plagioclase, amphibole, biotite, clinopyroxene, orthopyroxene, olivine, titanite Zircon, apatite, titanite KHM Plagioclase, biotite, amphibole, titanite Zircon, apatite, titanite High-K/shoshonite Andesite MB-B Clinopyroxene, feldspar, biotite, amphibole, quartz, orthopyroxene, olivine Zircon, apatite, titanite 57.8–58.5 3.5–4.0 MB-M Clinopyroxene, feldspar, biotite, amphibole, quartz, orthopyroxene, olivine Zircon, apatite, titanite 44 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Fig. 3. Selected zircon types from the studied samples (A, B, C: SEM-SE images of intact crystals; D: cathodoluminescence image of sectioned crystal). The scale is 100 μm. (A) slightly resorbed (Balvanyos); (B) resorbed (upper) and euhedral (lower) crystal (Baba-Laposa); (C) euhedral crystal (Dealul Mare); (D) typical internal structure of zircon with oscillatory zoning accompanied occasionally by sector zoning (Murgul Mare).

In most cases, the boundaries between the zones are sharp; how- values in excess of the accepted range of the distribution are ex- ever, rounded ones indicating resorption can be also observed around plained by simpliﬁed assumptions regarding zircon crystal geome- the interior domains in some crystals. These cores have typically higher try, heterogeneous distribution of parent nuclides, the presence of U concentrations (darker in CL images) and often contain thorite inclu- minute inclusions with elevated actinide contents and other imper- sions along the resorption boundary. The inclusion population in zircon fections of the crystals affecting the accuracy of the FT-correction crystals is quite uniform in all samples; apatite and glass inclusions are (e.g., Danišík et al., 2012). present in almost every crystal, whereas feldspar, amphibole, biotite, The obtained FT-corrected (U-Th)/He single-grain zircon dates Fe-Ti oxide and thorite inclusions are subordinate. are between 2.07 and 0.11 Ma (Table 4 and Appendix 2). Zircon crystals from the CAs samples are the oldest where the age range of the 4.3. Zircon (U-Th)/He data zircon dates varies between 2070 and 1760 ka (n = 6) and from 1700 to 1590 ka (n = 5) for the andesitic Murgul Mare and the dacite Single-grain (U-Th)/He zircon geochronology was applied to con- of the Pilişca volcano, respectively. The age range of the two HKSs strain the eruption ages of the peripheral lava domes of the CVDC samples is overlapping, their single-grain dates vary between 1050

(Table 4). Data having high analytical uncertainties (N10%), low FT and 920 ka (n = 8) and 1030 and 780 ka (n = 8) for the Malnaş values (b0.6) or unusually high U and Th content (implying the pres- and Bixad, respectively. The HKCAs cover a wider age interval com- ence of inclusions) were not considered in further calculations. The pared to the other two groups. The age range (1060–780 ka (n = distribution of the remaining dataset was tested for normality by 11)) of the Baba-Laposa which is oldest HKCAs sample overlaps using Q-Q-plots and Shapiro-Wilk test (Upton and Cook, 1996), with the HKSs samples. The single-grain dates of the andesitic Dealul which was conﬁrmed for all samples except for two outliers in sam- Mare and the dacitic Puturosul and Balvanyos vary 880–810 ka (n = ples KHM. A relatively small scatter of the data is indicated by the 6), 730–570 ka (n = 8) and 650–520 ka (n = 11), respectively. The mean square of weighted deviates (MSWD) of 0.1 to 2.0. MSWD two youngest samples among all of the dated units are the dacitic Turnul Apor and Haramul Mic, their age ranges vary between 340 and 310 ka (n = 3) and 170 and 110 ka (n = 12), respectively. Table 3 Zircon sizes and aspect ratios (minimum, maximum and average) of the studied samples. 4.4. The effect of secular disequilibrium Sample PD BL MB NH BH BAL AB KHM NM Length (μm) (U-Th)/He dating is based on the ingrowth of 4He under secular Minimum 326 133 147 121 149 136 114 162 125 equilibrium from the 238U, 235U, 232Th decay series and 147Sm decay Maximum 783 283 508 302 285 336 227 272 618 Average 438 204 243 200 203 238 165 202 219 (Farley, 2002). However, due to the fractionation of the intermediate 230 226 231 Aspect ratio daughter isotopes (e.g., Th, Ra, Pa), this condition is valid only Minimum 1.9 1.6 1.5 1.6 2.0 2.3 2.4 2.3 1.6 after ~5 half-lives of the longest-lived fractionating intermediate nu- Maximum 5.1 3.6 4.3 3.9 3.9 5.6 4.2 3.3 4.6 clide (Farley et al., 2002). This corresponds to N380 kyr prior to eruption Average 2.7 2.5 2.9 2.6 2.7 3.3 3.1 2.8 2.8 for the U-Th decay system and the expected fractionation of 230Th Table 4 Complete dataset of measurements, corrections, and calculations used to obtain ﬁnal (U-Th)/He eruption ages.

232 238 147 a 4 a b c d f g i i i Sample Sample Sample code Th ± U ± Sm ±% He ±% TAU Th/U Raw ±1 s.d. FT ± (%) FT-cor. ±1 s.d. D230 Disequilibrium +1σ −1σ ±1σ name ID (ng) %a (ng) %a (ng) (ncc) (%) (U-Th)/He (ka) (U-Th)/He (ka) corrected (ka) (ka) (ka) age (ka) agee (ka) (U-Th)/He ageh (ka)

Haramul KHM CSO-KHM-1 z1 2.406 2.4 2.819 1.8 0.003 25.1 0.043 3.0 3.4 0.85 105.5 3.5 0.727 8.2 145.1 12.9 0.263 178.9 16.7 20.2 18.4 Mic CSO-KHM-1 z2 2.639 2.4 3.979 1.8 0.001 37.6 0.055 2.9 3.3 0.66 99.4 3.3 0.777 6.7 127.8 9.5 0.205 154.3 17.6 13.1 15.4 CSO-KHM-1 z3 4.920 2.4 5.863 1.8 0.003 22.2 0.086 2.5 2.9 0.84 101.8 2.9 0.771 6.9 132.1 9.8 0.259 158.6 15.4 14.8 15.1 CSO-KHM-1 7.437 2.4 4.666 1.8 0.003 26.5 0.146 2.1 2.6 1.59 187.9 4.8 0.747 7.6 251.7 20.2 0.492 z4m CSO-KHM-1 z5 3.145 2.4 3.840 1.8 0.003 20.6 0.064 2.7 3.1 0.82 115.1 3.6 0.797 6.1 144.4 9.9 0.253 177.4 14.2 15.2 14.7 CSO-KHM-1 z6 5.841 2.4 5.890 1.8 0.008 16.3 0.115 2.2 2.7 0.99 130.8 3.5 0.814 5.6 160.7 10.0 0.306 194.3 12.0 13.2 12.6 CSO-KHM-1 1.478 2.4 1.643 1.8 0.006 24.3 0.034 3.2 3.5 0.90 142.1 5.0 0.747 7.6 190.2 15.9 0.278 z7m CSO-KHM-1 z8l 2.459 2.4 2.895 1.8 0.013 12.7 0.073 22.1 22.2 0.85 174.7 38.7 0.723 8.3 241.5 57.2 0.262 39 (2018) 354 Research Geothermal and Volcanology of Journal / al. et Molnár K. CSO-KHM-1 6.436 2.4 5.143 1.8 0.024 10.7 0.118 2.2 2.6 1.25 147.0 3.9 0.777 6.7 189.3 13.6 0.386 z9m CSO-KHM-1 2.691 2.4 3.839 1.8 0.011 15.8 0.053 2.6 3.0 0.70 97.8 2.9 0.747 7.6 131.0 10.7 0.216 160.1 17.6 16.5 17.1 z10 CSO-KHM-1 2.282 2.4 3.331 1.8 0.007 16.0 0.041 3.0 3.3 0.69 87.6 2.9 0.742 7.7 118.1 10.0 0.212 141.3 15.3 14.7 15.0 z11 CSO-KHM-1 1.863 2.4 2.692 1.8 0.012 14.3 0.037 3.3 3.7 0.69 98.9 3.6 0.705 8.8 140.3 13.4 0.214 176.4 18.7 22.0 20.3 z12 in_CSO-KHM-1 3.332 2.4 4.509 1.8 0.014 4.8 0.058 2.7 3.1 0.74 91.5 2.8 0.719 8.4 127.3 11.5 0.228 155.5 16.6 19.1 17.8 z12 in_CSO-KHM-1 1.739 2.4 2.711 1.8 0.007 4.8 0.031 3.5 3.8 0.64 83.3 3.2 0.696 9.1 119.6 11.8 0.198 144.6 18.2 17.9 18.0 z13 in_CSO-KHM-1 3.852 2.4 4.999 1.8 0.015 4.8 0.068 2.3 2.8 0.77 95.2 2.6 0.772 6.9 123.4 9.1 0.238 147.7 16.3 16.4 16.3 z5 in_CSO-KHM-1 6.470 2.4 5.247 1.8 0.031 4.8 0.082 2.4 2.8 1.23 100.2 2.8 0.762 7.1 131.6 10.1 0.381 151.0 15.3 16.6 15.9 z6 Eruption 163 (±11) MSWD g.o.f.: 0.321 age (±2σ ∗ = 1.4 √MSWD)j: Turnul AB CSO-AB z1 2.259 2.4 3.141 1.8 0.003 20.0 0.110 2.4 2.8 0.72 247.8 7.0 0.744 7.7 333.0 27.2 0.199 351.6 31.1 34.1 32.6 Apor CSO-AB z2 4.816 2.4 5.845 1.8 0.010 12.3 0.220 1.9 2.4 0.82 261.3 6.3 0.773 6.8 338.0 24.4 0.228 355.3 29.3 29.9 29.6 CSO-AB z3 3.825 2.4 4.969 1.8 0.011 9.6 0.174 1.9 2.4 0.77 245.9 6.0 0.779 6.6 315.6 22.3 0.213 330.4 24.8 27.1 26.0 Eruption 344 (±33) MSWD g.o.f.: 0.782 age (±2σ ∗ = 0.2 √MSWD)j: – Balvanyos BAL MK-2 z1 6.904 2.4 12.708 1.8 0.059 9.3 0.834 1.9 2.5 0.54 482.1 12.0 0.838 4.8 575.0 31.3 0.153 578.8 33.9 32.4 33.2 56 MK-2 z2 4.601 2.4 7.137 1.8 0.032 9.8 0.430 2.0 2.5 0.64 432.8 11.0 0.793 6.2 545.5 36.5 0.181 545.6 40.7 34.5 37.6 MK-2 z3 5.110 2.4 6.974 1.8 0.038 9.6 0.484 2.0 2.5 0.73 490.4 12.4 0.785 6.4 624.3 43.2 0.206 626.7 50.3 41.9 46.1 MK-2 z4 4.088 2.4 7.804 1.8 0.035 9.8 0.534 2.0 2.5 0.52 504.8 12.9 0.792 6.3 637.7 43.1 0.147 647.9 44.7 48.9 46.8 MK-2 z5 6.401 2.4 11.634 1.8 0.049 9.6 0.661 2.0 2.5 0.55 416.6 10.5 0.793 6.2 525.3 35.2 0.154 524.5 39.9 32.4 36.2 MK-2 z6 9.839 2.4 15.556 1.8 0.111 9.4 1.066 1.9 2.5 0.63 493.7 12.2 0.828 5.2 596.1 34.1 0.178 598.2 39.6 33.1 36.4 CSO-BAL(cs) z1 2.967 2.4 5.091 1.8 0.019 17.0 0.341 1.6 2.2 0.58 488.5 11.0 0.760 7.2 642.9 48.5 0.165 646.1 58.0 47.6 52.8 CSO-BAL(cs) z2 3.646 2.4 5.395 1.8 0.018 17.6 0.314 1.6 2.2 0.68 416.2 9.2 0.818 5.5 509.0 30.1 0.191 508.8 33.5 28.6 31.0 CSO-BAL(cs) z3 2.545 2.4 5.337 1.8 0.012 20.1 0.375 1.6 2.2 0.48 522.6 11.7 0.807 5.8 647.3 40.1 0.135 651.6 50.1 39.3 44.7 CSO-BAL(cs) z4 3.998 2.4 5.685 1.8 0.030 10.9 0.372 1.6 2.2 0.70 465.1 10.2 0.783 6.5 594.0 40.8 0.199 595.4 46.3 39.5 42.9 CSO-BAL(cs) z5 6.913 2.4 9.067 1.8 0.061 10.8 0.624 1.5 2.1 0.76 483.4 10.3 0.825 5.3 586.2 33.3 0.216 587.1 38.7 31.4 35.0 Eruption 583 (±30) MSWD g.o.f.: 0.097 age (±2σ ∗ = 1.6 √MSWD)j: Puturosul BH CSO-BH z1 2.389 2.4 2.067 1.8 0.001 34.7 0.167 2.0 2.5 1.16 526.9 13.3 0.728 8.1 723.3 61.7 0.301 741.7 73.5 73.2 73.4

(continued on next page) 45 46 Table 4 (continued)

CSO-BH z2 4.718 2.4 4.548 1.8 0.031 9.0 0.355 1.8 2.3 1.04 518.4 12.1 0.792 6.2 654.4 43.6 0.270 659.7 50.9 44.3 47.6 CSO-BH z3 2.893 2.4 4.335 1.8 0.003 29.5 0.287 1.8 2.4 0.67 472.9 11.1 0.820 5.4 576.8 34.0 0.174 579.9 38.2 33.6 35.9 CSO-BH z4 3.640 2.4 3.678 1.8 0.004 26.0 0.291 1.8 2.3 0.99 531.6 12.4 0.744 7.7 714.4 57.3 0.258 732.8 69.9 68.6 69.2 CSO-BH z5l 3.693 2.4 5.006 1.8 0.005 17.1 0.358 1.8 2.4 0.74 504.6 11.9 0.554 13.4 911.4 123.9 0.192 CSO-BH z6 1.853 2.4 2.983 1.8 0.002 32.7 0.210 1.9 2.4 0.62 509.7 12.5 0.739 7.8 689.4 56.5 0.162 699.6 74.9 60.4 67.7 CSO-BH(sz) z1 2.419 2.4 3.268 1.8 0.017 16.2 0.211 2.0 2.5 0.74 455.1 11.3 0.777 6.7 585.4 41.7 0.200 586.5 47.7 40.2 44.0 CSO-BH(sz) z2 1.463 2.4 2.380 1.8 0.012 18.3 0.163 2.1 2.6 0.61 495.9 12.9 0.754 7.4 657.7 51.4 0.166 663.0 63.9 52.2 58.0 CSO-BH(sz) z3 1.282 2.4 4.028 1.8 0.015 18.7 0.269 1.9 2.5 0.32 514.4 12.7 0.761 7.2 675.6 51.2 0.086 686.3 67.9 54.8 61.3 Eruption 642 (±44) MSWD g.o.f.: 0.219 age (±2σ ∗ = 1.4 √MSWD)j: 39 (2018) 354 Research Geothermal and Volcanology of Journal / al. et Molnár K. Dealul NH CSO-NH5 z1 3.183 2.4 3.470 1.8 0.007 15.1 0.332 1.7 2.3 0.92 651.4 14.8 0.741 7.8 878.5 71.0 0.285 885.4 99.1 74.8 86.9 Mare CSO-NH5 z2 2.580 2.4 4.268 1.8 0.008 14.5 0.387 1.7 2.3 0.60 657.8 15.1 0.801 6.0 821.1 52.5 0.188 828.4 58.7 55.2 57.0 CSO-NH5 z3 1.959 2.4 3.085 1.8 0.007 15.8 0.261 1.9 2.5 0.64 610.5 15.1 0.748 7.6 816.3 65.0 0.198 820.5 71.2 66.0 68.6 CSO-NH5 z4 12.951 2.4 7.076 1.8 0.016 11.0 0.782 1.6 2.2 1.83 638.9 13.8 0.783 6.5 816.1 56.0 0.569 815.1 63.0 53.3 58.2 CSO-NH5 z5 3.779 2.4 6.851 1.8 0.005 19.8 0.678 1.6 2.2 0.55 725.2 16.2 0.830 5.1 873.5 48.6 0.172 888.1 68.1 54.9 61.5 CSO-NH5 z6 1.715 2.4 2.835 1.8 0.003 28.7 0.250 1.8 2.4 0.61 639.6 15.4 0.771 6.9 829.5 60.4 0.188 836.4 69.2 63.5 66.4 Eruption 842 (±53) MSWD g.o.f.: 0.927 age (±2σ ∗ = 0.2 √MSWD)j: Bixad MB-B CSO-MB-B z1 3.416 2.4 5.377 1.8 0.028 12.4 0.479 1.6 2.2 0.64 641.3 14.1 0.819 5.4 783.4 45.9 0.167 788.9 42.2 51.6 46.9 CSO-MB-B z2 5.790 2.4 10.203 1.8 0.066 9.4 0.982 1.4 2.1 0.57 703.0 14.9 0.835 4.9 841.9 45.3 0.150 843.3 47.2 46.5 46.8 CSO-MB-B z3 7.658 2.4 9.062 1.8 0.074 11.5 1.019 1.4 2.1 0.85 776.4 15.9 0.823 5.3 943.6 53.8 0.223 942.9 60.3 51.2 55.8 CSO-MB-B z4 3.463 2.4 6.325 1.8 0.028 13.7 0.617 1.5 2.1 0.55 715.2 15.3 0.807 5.8 886.8 54.9 0.144 896.2 49.9 64.0 57.0 CSO-MB-B z5 3.552 2.4 4.404 1.8 0.032 12.5 0.503 1.5 2.1 0.81 793.9 16.8 0.747 7.6 1062.1 83.6 0.213 1067.4 98.4 87.4 92.9 CSO-MB-B z6 3.216 2.4 5.052 1.8 0.014 18.2 0.503 1.5 2.2 0.64 717.1 15.4 0.742 7.7 966.7 77.7 0.168 964.7 87.3 76.1 81.7 MB z8 2.720 2.4 4.162 1.8 0.063 24.9 0.476 1.2 2.0 0.65 821.1 16.2 0.800 6.0 1026.8 64.9 0.172 1030.1 72.7 65.9 69.3 MB z9 4.309 2.4 6.814 1.8 0.165 20.0 0.756 1.1 1.9 0.63 799.2 15.2 0.799 6.0 1000.8 63.4 0.167 1003.6 70.3 63.6 67.0 Eruption 907 (±66) MSWD g.o.f.: 0.006 age (±2σ ∗ = 2.5 √MSWD)j: Baba BL CSO-BL-1 z1l 23.819 2.4 15.492 1.8 0.037 6.8 1.495 1.5 2.1 1.54 585.8 12.1 0.807 5.8 726.1 44.7 0.464 Laposa CSO-BL-1 z2 4.143 2.4 5.630 1.8 0.007 16.8 0.603 1.6 2.2 0.74 756.1 16.7 0.805 5.9 939.5 58.8 0.222 942.8 66.2 58.5 62.4 CSO-BL-1 z3 15.021 2.4 8.756 1.8 0.090 6.0 1.247 1.5 2.1 1.72 838.9 17.4 0.805 5.8 1041.9 64.6 0.517 1047.8 83.9 63.4 73.7 CSO-BL-1 z4 1.797 2.4 2.769 1.8 0.002 34.3 0.293 1.9 2.4 0.65 760.0 18.3 0.773 6.8 983.4 71.1 0.196 989.8 94.3 74.6 84.5 CSO-BL-1 z5 2.106 2.4 3.043 1.8 0.015 10.9 0.347 1.7 2.3 0.69 812.6 18.8 0.769 6.9 1056.5 77.1 0.209 1101.2 79.5 109.5 94.5 – CSO-BL-1 z6 2.637 2.4 3.527 1.8 0.002 40.1 0.379 1.7 2.3 0.75 757.4 17.1 0.774 6.8 978.9 70.0 0.225 986.7 85.4 75.7 80.5 56 CSO-BL-1 zR2 2.813 2.4 4.046 1.8 0.041 15.9 0.417 1.8 2.4 0.70 733.2 116.9 0.777 6.7 943.4 67.0 0.210 949.0 73.4 71.0 72.2 CSO-BL-1 zR4 5.016 2.4 4.437 1.8 0.034 13.8 0.425 1.9 2.4 1.13 625.7 86.1 0.800 6.0 782.1 50.5 0.341 779.7 55.2 47.3 51.3 CSO-BL-1 zR5 1.147 2.4 1.885 1.8 0.007 37.6 0.202 2.0 2.5 0.61 774.4 290.9 0.740 7.8 1046.2 85.7 0.183 1107.4 73.7 140.4 107.1 CSO-BL-1 zR6 1.703 2.4 1.964 1.8 0.000 0.0 0.193 2.0 3.6 0.87 674.6 0.0 0.730 8.1 924.3 82.1 0.261 925.5 87.3 82.6 85.0 CSO-BL-1 zR7 4.824 2.4 6.248 1.8 0.030 21.7 0.606 1.8 2.3 0.77 679.8 147.6 0.793 6.2 857.5 57.0 0.233 858.1 63.2 56.9 60.1 CSO-BL-1 zR8 4.216 2.4 4.130 1.8 0.017 24.2 0.460 1.8 2.4 1.02 743.1 179.8 0.709 8.7 1047.4 94.6 0.308 1093.3 89.5 139.0 114.3 Eruption 942 (±65) MSWD g.o.f.: 0.044 age (±2σ ∗ = 2.2 √MSWD)j: Malnaş MB-M CSO-MB-M z4 2.687 2.4 3.082 1.8 0.016 13.6 0.355 1.2 1.9 0.87 791.3 15.3 0.754 7.4 1049.2 80.0 0.235 1047.1 87.0 77.5 82.3 CSO-MB-M z5 4.460 2.4 5.782 1.8 0.053 7.5 0.651 1.3 2.0 0.77 789.3 15.7 0.816 5.5 966.9 56.6 0.208 963.0 63.8 52.5 58.2 CSO-MB-M z6 2.892 2.4 3.626 1.8 0.031 9.7 0.375 1.3 2.0 0.80 719.9 14.4 0.781 6.6 921.4 63.2 0.215 917.1 69.9 58.5 64.2 CSO-MB-M z7 2.318 2.4 3.603 1.8 0.013 16.5 0.387 1.3 2.0 0.64 773.0 15.5 0.777 6.7 994.8 69.5 0.174 992.1 76.3 65.7 71.0 CSO-MB-M z8 7.536 2.4 9.875 1.8 0.055 8.8 1.073 1.0 1.8 0.76 762.3 14.0 0.817 5.5 933.5 54.2 0.206 940.1 50.7 60.9 55.8 CSO-MB-M z9 3.426 2.4 4.902 1.8 0.034 8.1 0.505 1.2 1.9 0.70 732.2 14.2 0.781 6.6 937.6 64.3 0.189 944.7 60.2 71.7 65.9 CSO-MB-M z10 3.827 2.4 5.252 1.8 0.018 14.3 0.544 1.1 1.9 0.73 732.4 13.8 0.776 6.7 944.3 66.0 0.197 938.6 74.2 60.3 67.3 CSO-MB-M z11 2.612 2.4 3.319 1.8 0.014 15.0 0.380 1.2 1.9 0.79 799.7 15.5 0.773 6.8 1033.9 73.1 0.212 1031.8 80.3 70.6 75.4 Eruption 964 (±46) MSWD g.o.f.: 0.871 age (±2σ ∗ = 0.4 √MSWD)j: Pilişca PD CSO-PD2 z4 3.253 2.4 4.660 1.8 0.027 12.4 0.802 1.4 2.1 0.70 1223.6 25.8 0.766 7.0 1597.1 117.0 0.216 CSO-PD2 z5 2.114 2.4 3.413 1.8 0.050 12.0 0.573 1.5 2.1 0.62 1213.8 26.1 0.758 7.3 1601.5 121.3 0.192 CSO-PD2 z6 4.481 2.4 4.929 1.8 0.046 9.5 0.877 1.4 2.1 0.91 1212.6 25.1 0.735 7.9 1649.5 135.5 0.281 CSO-PD2 z7 2.661 2.4 4.720 1.8 0.087 9.1 0.857 1.4 2.1 0.56 1327.4 27.8 0.782 6.5 1696.7 116.4 0.174 CSO-PD2 z8 2.040 2.4 2.551 1.8 0.032 11.4 0.436 1.5 2.1 0.80 1189.3 25.3 0.714 8.6 1664.5 146.9 0.247 Eruption 1640 MSWD = age (±37) 0.1 (±2σ)k: Murgul NM CSO-NM1 z1 0.844 2.4 2.352 1.8 0.011 19.3 0.444 1.2 2.0 0.36 1443.1 29.2 0.777 6.7 1856.9 129.7 0.100 Mare CSO-NM1 z2 1.666 2.4 3.818 1.8 0.013 14.5 0.787 1.1 1.9 0.44 1547.7 30.1 0.795 6.2 1947.8 125.9 0.121 CSO-NM1 z3 1.154 2.4 3.029 1.8 0.011 18.9 0.576 1.2 2.0 0.38 1444.8 28.7 0.766 7.0 1886.6 137.7 0.106 CSO-NM1 z4 2.894 2.4 7.253 1.8 0.022 13.3 1.389 1.0 1.9 0.40 1450.2 27.5 0.823 5.3 1761.5 99.2 0.111

CSO-NM1 z5 1.398 2.4 3.274 1.8 0.013 15.8 0.675 1.1 2.0 0.43 1551.4 30.3 0.750 7.5 2068.1 160.2 0.119 39 (2018) 354 Research Geothermal and Volcanology of Journal / al. et Molnár K. CSO-NM1 z6 0.814 2.4 2.056 1.8 0.009 19.4 0.373 1.3 2.0 0.40 1373.2 28.1 0.760 7.2 1807.9 135.5 0.110 Eruption 1865 MSWD = age (±87) 0.7 (±2σ)k:

Note: σ = standard error, s.d. = standard deviation, MSWD = mean square of weighted deviates, g.o.f. = goodness of fit. a Analytical uncertainties associated with the measurement of each isotope. b TAU = total analytical uncertainty. c Ft is alpha ejection correction factor calculated after Farley et al. (1996), Farley (2002) and Hourigan et al. (2005). d Calculated by 30 ∗ (1 − Ft). e Ft-corrected (α-ejection corrected) (U-Th)/He age. f Uncertainty including analytical uncertainties and alpha ejection correction uncertainties. g Zircon/magma D230 parameters were calculated using the Th/U ratios of analyzed bulk zircon crystals and whole rock (Farley et al., 2002). h Disequilibrium corrected (U-Th)/He age calculated by MCHeCalc program (Schmittetal.,2010). i Uncertainty includes previous propagated errors and disequilibrium correction as calculated by MCHeCalc program (Schmittetal.,2010). j “Concordant” eruption age calculated by MCHeCalc program (Schmittetal.,2010). k Weighted mean age. l Outliers omitted from the final age calculation because of their high TAU, low FT or unusually high U and Th content. m Outliers omitted from the final age calculation because of the distribution test (Q-Q plot, Shapiro-Wilk test, Upton and Cook, 1996). – 56 47 48 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 relative to parental 234U (reasonably expected to be in equilibrium with different corrections on the accuracy of the final date. Zircon/magma 238 U). The effect of isotope disequilibrium has to be considered, how- D230 parameters (Farley et al., 2002) were calculated using the Th/U ra- ever, only for Quaternary samples, because for older zircons the un- tios of analyzed bulk zircon crystals and whole rock (Table 4). We as- 231 4 certainty related to the (U-Th)/He-method is greater than the sumed secular equilibrium for Pa (i.e. D231 =1)becausethe He secular disequilibrium effect on the ages (Farley et al., 2002). This contribution from the 235U decay series is minor, and the effects of correction can be performed based on the crystallization ages of an- disequilibrium are negligible within reasonable limits of D231 alyzed zircon crystals obtained by U-Th or U-Pb zircon dating (Schmitt, 2011). The results of these calculations are presented in (e.g., Schmitt et al., 2006, 2010; Schmitt, 2011; Danišík et al., Table 4, Appendix 2 and Fig. 4. Uncertainties were calculated by mul- 2012, 2017; Lindsay et al., 2013). If crystallization ages cannot be tiplying the 2-sigma error with the square root of the MSWD obtained, or are heterogeneous, the “secular equilibrium assumed” (e.g., Danišík et al., 2012). (U-Th)/He age and the “full disequilibrium assumed” (U-Th)/He The “full disequilibrium assumed” (U-Th)/He time yields the maxi- age will give the upper and lower age limits for the eruption age, re- mum possible eruption age for every sample, assuming no zircon resi- spectively. Full disequilibrium zircon (U-Th)/He ages mean that zir- dence time (Appendix 2). However, it is very useful also for con crystallized just before eruption (i.e., the zircon crystallization identifying the outlier dates and indicating the necessity of the Th- time is equal to the eruption age; Farley et al., 2002), whereas the disequilibrium correction (i.e., if the “secular equilibrium assumed” “secular equilibrium assumed” (U-Th)/He age scenario applies to and the “full disequilibrium assumed” ages are overlapping with each crystals with protracted (N375 kyr) pre-eruptive storage or to other, no Th-disequilibrium correction is needed). The two oldest CAs- xenocrysts. samples yielded (U-Th)/He weighted mean ages of 1954 ± 48 ka The studied samples yielded (U-Th)/He ages b2 Ma and therefore, (Murgul Mare; NM) and 1696 ± 22 ka (Pilişca; PD) for “full disequilib- the effects of isotope disequilibrium should be considered. For this rium assumed”, which are indistinguishable within uncertainties from purpose, Schmitt et al. (2006) used a combined U-Th and (U-Th)/He the disequilibrium-uncorrected He ages (1865 ± 87 ka and 1640 ± analysis, when the Ft-corrected (U-Th)/He data were corrected by the 37 ka, respectively). Therefore, no disequilibrium correction is required average crystallization age of the zircons obtained from U-Th spot for these samples and the “secular equilibrium assumed” (U-Th)/He interior ages. Later, Schmitt et al. (2010) refined this method and ages approximate well the eruption ages. For all the other samples measured the U-Th and (U-Th)/He dates on the same zircon grains with ages b1.5 Ma, the “full disequilibrium assumed” (U-Th)/He ages (double-dating). Although Danišík et al. (2017) suggested that zircon (Appendix 2) differ from the “secular equilibrium assumed” (U-Th)/ double-dating is the most accurate way to determine the eruption He ages, thus a disequilibrium correction is applied (Table 4). The cor- ages, the two methodologies provide similar result within uncertainty rection significantly increases zircon dates for samples b500 ka, (e.g., Schmitt et al., 2010; Danišík et al., 2012). Schmittetal.(2010)em- whereas the U-Pb corrected (U-Th)/He ages for samples with ages be- phasized that in case of age-zoned crystals, the rim age provides a lower tween ca. 1 Ma and 500 ka shifted only slightly because of the large limit for the bulk crystallization age and the eruption age would be (i.e., ~100 kyr) average zircon residence time during which zircon overcorrected. Zircons in the Ciomadul dacites show a wide range of largely attained secular equilibrium prior to eruption. crystallization age, even in single crystals (Harangi et al., 2015a), there- A direct comparison of the double-dating method (Schmitt et al., fore, we used the weighted average U-Th (for the sample KHM) and U- 2010; Danišík et al., 2012, 2017), and the average crystallization age Pb (for all the other samples) dates (Table 5) reported in Lukács et al. method (Schmitt et al., 2006; Danišík et al., 2012; Harangi et al., 2015a) (submitted) to correct the initial Th disequilibrium. The corrected ages was carried out for sample KHM (Table 6). The combined U-Th and (U- were computed using the MCHeCalc software (http://sims.ess.ucla. Th)/He double-dating results are 154 ± 16 ka (n =4,goodnessoffit: edu/Research/MCHeCalc.php) following the description provided by 0.952), which is slightly younger but overlaps within uncertainty with Schmitt et al. (2010). However, in case of the youngest sample (KHM) the average crystallization age corrected (U-Th)/He age (i.e., 163 ± both double-dating and average crystallization age correction were 11 ka; Fig. 5). Both calculated eruption ages are consistent with the youn- applied (1) to perform accurate eruption age with the standard gest rim age of the zircon grains (142 ± 18 ka; Table 5; Lukács et al., technique (Schmitt et al., 2006, 2010) and (2) to monitor the effect of submitted), which gives a maximum age limit for the eruptions (Fig. 4).

Table 5 The average U-Pb and U-Th crystallization ages of the studied samples calculated from data presented by Lukács et al. (submitted).

Sample Sample ID Method Number of Weighted Uncertainty (2σ; LA-ICP-MS MSWD Youngest Uncertainty (2σ; LA-ICP-MS name data mean and spot and (used/all) age [ka] 1σ**; SIMS) [ka] age [ka] 1σ**; SIMS) [ka]

Haramul CSO-KHM1 SIMS mantlea U-Th 24/25 198 15** 2.3 154 14** Mic CSO-KHM1 SIMS rim U-Th 21/21 179 11** 1.1 142 18** Turnul Apor CSO-AB LA-ICP-MS U-Pb 33/49 471 35 28 395 29 mantlea CSO-AB SIMS rim 20/20 365 47** 0.37 304 96** Balvanyos MK2 SIMS mantlea 18 864 68** 1.6 608 99** CSO-BAL(cs) SIMS rim 20/20 731 60** 2.1 577 74** CSO-BAL03 LA-ICP-MS 20/20 803 58 6.4 658 80 rim Puturosul CSO-BH LA-ICP-MS 20/23 841 47 19 730 46 mantlea CSO-BH SIMS rim 5/5 715 87** 2.1 650 91** Dealul Mare CSO-NH5a 40/44 1004 33 11.2 864 60 Bixad CSO-MB-Ba 33/38 1226 38 10.8 1075 62 Baba-Laposa CSO-BL1a 38/44 1129 34 8.3 980 72 Malnaş CSO-MB-Ma 38/43 1332 47 25 1140 92 Pilişca CSO-PD2 40/45 1867 42 1.3 1708 85 Murgul CSO-NM1 47/60 2165 45 9.3 1940 95 Mare

a Average U-Th and U-Pb crystallization ages which were used for the disequilibrium correction in MCHeCalc program. K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 49

Fig. 4. Combined U-Th-Pb and (U-Th)/He zircon ages. Each panel shows the “secular equilibrium-assumed” (rectangle), the disequilibrium-corrected (circle) and “full disequilibrium- assumed” (triangle) (U-Th)/He ages, the weighted means of the disequilibrium corrected ages (filled circle/rectangle), a Gaussian fit to replicate analyses representing the modeled “concordant” eruption ages (red Gaussian curve) and the youngest crystallization ages (filled star) with uncertainties. The obtained eruption ages with 2σ ∗√MSWD uncertainties are indicated in every panel for each sample. Color coding of panels/samples see Fig. 2.

5. Discussion ages of the youngest volcanic phase of the Ciomadul volcano using the same methodology as in this study. The eruption ages derived from 5.1. Eruption chronology the zircon (U-Th)/He dates corrected for disequilibrium correspond to 56 to 32 ka. This dominantly explosive volcanic episode was preceded The Ciomadul Volcanic Dome Complex is the youngest manifesta- by an effusive phase with prevailing lava dome extrusions, where the tion of the Neogene to Quaternary volcanism of the entire Carpathian- uncorrected zircon (U-Th)/He dates presented in Karátson et al. Pannonian region. Harangi et al. (2015a) determined the eruption (2013) indicate eruption ages between ca. 200 and 100 ka. Here, the

Table 6 Dataset of measurements and corrections of the double dated KHM sample. For the detailed dataset of the (U-Th)/He analysis and the footnotes please refer to Table 4.

238 232 230 232 g i i i Sample Sample Sample code ( U)/( Th) ± ( Th)/( Th) ± FT-cor. ±1 D230 U-Th + − Disequilibrium +1σ −1σ ±1σ name ID (U-Th)/He s.d.f age corrected (ka) (ka) (ka) (ka) agee (ka) (ka) (ka) (U-Th)/He ageh (ka)

Haramul KHM in_CSO-KHM-1 8.44 0.2 6.73 0.2 127.3 11.5 0.228 162 19 16 155.5 16.6 19.1 17.8 Mic z12 in_CSO-KHM-1 8.41 0.2 7.26 0.2 119.6 11.8 0.198 172 22 23 144.6 18.2 17.9 18.0 z13 in_CSO-KHM-1 6.52 0.2 5.44 0.2 123.4 9.1 0.238 181 29 23 147.7 16.3 16.4 16.3 z5 in_CSO-KHM-1 3.44 0.1 2.95 0.1 131.6 10.1 0.381 180 36 27 151.0 15.3 16.6 15.9 z6 Eruption 154 (±16) g.o.f.: 0.952 age (±2σ)j: 50 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

and robust. Furthermore, the concordant ages of radiocarbon, thermolu- minescence, and (U-Th)/He zircon ages for the same eruptive products of the youngest volcanic rocks of the Ciomadul volcano (Harangi et al., 2015a) validate the applicability and accuracy of the (U-Th)/He zircon geochronology on these young volcanic rocks. Palaeomagnetic data of Panaiotu et al. (2012) serve as an independent control for the proposed eruption ages. Our new combined U-Th- Pb and (U-Th)/He zircon ages are in good agreement with published palaeomagnetic polarity data (Fig. 6). Specifically, normally magnetized Malnaş could correspond to the short-lived Santa Rosa excursion, which is possibly also the case for Bixad, although eruption during the Kamikatsura excursion (Singer, 2014) is also conceivable within analytical uncertainties. The normally magnetized Balvanyos and the reversed Pilişca correspond to the Bruhens and Matuyama chrons, respectively. The eruption age of Murgul Mare which has a transitional magnetic character (Panaiotu et al., 2012) correlates well with the Olduvai excursion. This new, detailed dating of the eruption products of the older phase of the CVDC enables to reconstruct the following eruption history for the South Harghita over the past 2 Myr (Fig. 7). The oldest two samples are the 1865 ± 87 ka andesitic lava dome of Murgul Mare and the younger, 1640 ± 37 ka dacitic part of the Pilişca volcano, which predate well the activity of the CVDC (Fig. 7A). Further eruptions during this period cannot be excluded, although, the PD dacite is inferred to be one of the youngest volcanic product at Pilişca (Szakács et al., 2015). In this case, a ca. 600 kyr quiescence period Fig. 5. Comparison between the double-dating (a) and the average crystallization age preceded the initial CVDC volcanism. Within a relatively short period correction (b) method. Uncertainties for the obtained ages are 2σ for the double-dating b σ ∗√ ( 100 kyr), three lava domes were formed: two HKS lava bodies approach (a) and 2 MSWD for the average crystallization age correction (b). U-Th ş ages of the dated zircon grains (star), youngest crystallization age (filled star) and (Malna : 964 ± 46 ka; and Bixad: 907 ± 66 ka) in the vicinity to average crystallization age (grey bar) with uncertainties are indicated, further symbols the site of the Murgul Mare and the first HKCA lava dome of Baba- and colors as in Fig. 4. Laposa (942 ± 65 ka), which emerged at the foot of the Pilişca volcano. In particular, extrusion of the Baba-Laposa dacite with the unique chemical composition, which is indistinguishable from the eruption chronology is extended to the initial stage of CVDC volcanism, later CVDC dacites could be a sign of the onset of the development when multiple isolated, mainly dacitic lava domes were formed. We of the CVDC. These three lava domes are arranged parallel to the dated all identified localities, where K/Ar ages were already determined river Olt (Fig. 7B). This alignment corresponds to a NNE-SSW (Table 1.), but additionally the Turnul Apor lava was also sampled trending normal fault system, which follows the orientation of the which has not been dated yet. This complete eruption age coverage main normal faults of the Braşov (and Ciuc) basin (Girbacea et al., for the early CVDC yields a precise age framework on the repose time 1998; Ciulavu et al., 2000; Seghedi et al., 2011) and is perpendicular between eruptions. (i.e., NNE-SSW) to the NW-SE orientation of the CGH volcanic chain. Eruption ages inferred from (U-Th)/He zircon cooling ages (Table 4, After a short (b100 kyr) repose time, an andesitic HKCA lava dome Appendix 2) for the peripheral CVDC lava domes are in the range be- (Dealul Mare: 842 ± 53 ka) was formed which was followed by two tween 964 and 580 ka (the 344 ka Turnul Apor is not in a peripheral po- adjacent dacitic lava domes (Puturosul: 642 ± 44 ka and Balvanyos: sition), whereas the Haramul Mic lava dome was erupted at 154 ± 583 ± 30 ka; Fig. 7B) after ca. 100 kyr quiescence. After another, ca. 16 ka. These ages, which are supported also by the youngest zircon crys- 100 kyr dormancy the Turnul Apor dacitic lava (344 ± 33 ka) erupted. tallization ages (Table 5; Lukács et al., submitted) are much younger The four lava domes of Dealul Mare, Puturosul, Balvanyos and Turnul than the previously determined K/Ar ages (Pécskay et al., 1992, 1995; Apor are situated on the eastern side of the river Olt and their positions Szakács et al., 2015; Table 1). The most pronounced difference is follow the NW-SE orientation of the CGH volcanic chain. The central found between the K/Ar age and zircon (U-Th)/He age of the Haramul lava dome complex of the Ciomadul volcano is located perpendicular Mic dacite (850 ka in Casta (1980), while 154 ka in this study) and the to the NW-SE orientation of the CGH volcanic chain and shows a similar Malnaş-Bixad shoshonites (2.2–1.4 Ma in Peltz et al. (1987), while elongation as the NNE-SSW aligned ca. 1000–900 ka Malnaş, Bixad and 964–907 ka in this study). Because most of the studied sites (excluding Baba-Laposa lava domes. The onset of volcanism at the more volumi- the Pilişca volcanic complex) represent small-volume lava domes nous (~8–9km3; based on DEM analysis Szakács et al., 2015) (~0.1–0.6 km3 based on DEM analysis; Szakács et al., 2015), they likely Ciomadul volcano occurred b200 ka (based on the uncorrected He formed monogenetically during a short period of time. Evidence for ages of the oldest lava dome rocks presented in Karátson et al., eruptive hiatuses is absent, and it is thus unlikely that the differences 2013), i.e., after another protracted (N100 kyr) quiescence period. between K/Ar and (U-Th)/He dates are due to age heterogeneity within The Haramul Mic lava dome is separated from the main massif of individual domes. Instead, it is conceivable that K/Ar dates are affected the volcano, but its age (154 ± 16 ka) clearly suggests that it belongs by unidentified excess argon, and therefore yield apparent older ages to the lava dome building phase of Ciomadul volcano (Fig. 7C). A (e.g., Hora et al., 2007; Hildenbrand et al., 2012). The contribution of ex- prolonged quiescence period repeats between the Ciomadul lava cess 40Ar is potentially aggravated by the abundance of porphyritic dome extrusion phase and the more dominant explosive volcanic rocks in the CVDC, where mineral phases crystallized during open- phase between around 100 and 60 ka. system magmatic processes over a prolonged period in a crustal In summary, long repose times between eruptions characterize the magma reservoir (Kiss et al., 2014; Harangi et al., 2015a, 2015b). The volcanism of the CVDC. Such long quiescence between eruption phases unknown effect of excess argon during K/Ar dating makes zircon- is not uncommon in volcanoes fed by intermediate magmas (e.g., Lassen based geochronology – in this case (U-Th)/He dating – more reliable Peak, USA, Clynne and Muffler, 2010; Aucanquilcha, Andes, Chile; K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 51

Fig. 6. Temporal evolution of the studied samples supplemented with uncorrected He ages published in Karátson et al. (2013), combined U-Th and (U-Th)/He ages from Harangi et al. (2015a) and the available polarity information (indicated by (N: normal), (R: reversed) or (T: transitional)) from Panaiotu et al. (2012). Geomagnetic Polarity Time Scale is after Singer (2014). CV: Ciomadul volcano. Color coding of samples as in Fig. 2.

Fig. 7. A–C: Volcanological evolution of the South Harghita in the past 2 million years based on the eruption ages calculated from zircon (U-Th)/He dates (Harangi et al., 2015a; this study) and the disequilibrium uncorrected He dates of the oldest lava dome rocks of the Ciomadul volcano (Karátson et al., 2013). Green and red lines indicate the coeval volcanic activity at ca. 1 Ma and 150–100 ka, respectively. The eruption centers in these two time intervals (B and C) are arranged perpendicular to the normal, NW-SE trend of the CGH. D: Magma type distributions of the Southernmost Harghita. 52 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Klemetti and Grunder, 2008; Sabalan, Iran, Ghalamghash et al., 2016; system processes, i.e., mixing of distinct magmas (Mason et al., Ruapehu, New Zealand, Gamble et al., 2003; among others). The seem- 1995, 1996). After a short time (a few 10s kyr), another viscous, ingly inactive state, however, does not indicate that the magmatic crystal-rich magma batch with andesitic bulk composition extruded system beneath the volcano is inactive. Harangi et al. (2015a) forming the Dealul Mare lava dome. Although it is less silicic than the showed that zircon crystallization ages are in the range from ca. dominant HKCAs rocks of Ciomadul, it shows a clear geochemical 350 ka to 80 ka in the youngest volcanic products of Ciomadul sug- and petrologic affinity to them indicating possibly more mafic gesting a prolonged presence of melt-bearing magma. This implies magma component. After the eruption of the 842 ka Dealul Mare an- that the comparatively brief ca. 32 kyr quiescence after the most re- desite, there has been no major change in the composition of the cent CVDC eruption alone is not necessarily indicative for the mag- erupted magmas: they comprise homogeneous HKCAs dacite which matic system being already extinct. Multiple lines of evidence built up the more voluminous Ciomadul volcano. The compositional (Vaselli et al., 2002; Popa et al., 2012; Kiss et al., 2014; Harangi and textural variations of phenocrysts reflect still open-system et al., 2015a, 2015b; Kis et al., 2017) suggests continued presence magma genesis involving both maficandsilicicmagmas(Kiss et al., of a melt-bearing magma body beneath Ciomadul with a potential 2014). for rapid rejuvenation of this crystal mush. This led Harangi et al. (2015a, 2015b) to introduce a new term for such long-period repose 6. Conclusions volcanoes: volcanoes with Potentially Active Magma Storage (PAMS volcanoes). 1. The newly determined eruption ages enable us to constrain the onset of volcanism in the Ciomadul Volcanic Dome Complex, 5.2. Relations between magma type and eruption age the youngest volcanic area of eastern-central Europe. The eruption chronology has been significantly refined, revealing that The determined eruption ages along with the bulk rock composi- many eruptions are younger than previously thought based on tion of the dated volcanic products enable to delineate the changes in K/Ar data. The CVDC volcanism started at ca. 1 Ma after several the erupted magma composition. The first major change in magma 100s kyr of quiescence and was characterized by intermittent type occurred after the time gap at 3.9–2.8 Ma within the South lava dome extrusions until 344 ka. This was followed by the de- Harghita that divides the Luci-Lazu volcanic complex (4.3–3.6 Ma; velopment of the voluminous Ciomadul volcano from about 170 Pécskay et al., 1995) from the volcanoes located southward to 32 ka after another prolonged repose time. (i.e., Cucu, Pilişca and Ciomadul). The erupted magmas became 2. In the South Harghita region, at least three major erupted magma more potassic and show a different trace element signature than types can be distinguished. A sharp change in the erupted magma the older rocks to the north (Seghedi et al., 1987; Szakács et al., composition is recognized at ca. 1 Ma following an earlier composi- 1993; Mason et al., 1996; Harangi and Lenkey, 2007; Harangi et al., tional shift at 2.8 Ma as defined by Pécskay et al. (2006). Initially, dis- 2007; Vinkler et al., 2007). This change was interpreted as the result tinct high-K magma types erupted in the CVDC within a short (a few from (1) the tectonic inversion in the East Carpathians recorded at 10s kyr) period and in a spatially restricted area. However, since the beginning of the Quaternary and (2) the change in the collision 650 ka, volcanism is characterized by eruption of homogeneous mechanism caused by the difference in rheology of the continental dacitic magma. blocks, i.e., the East European Plate and the Moesian Plate north 3. Dating the volcanic activity of all identified eruptive CVDC cen- and south from the Trotuş fault (Cloetingh et al., 2004; Harangi and ters, predating the development of the Ciomadul volcano, Lenkey, 2007; Seghedi et al., 2011). Although, the post-2.8 Ma allowed us to constrain the repose intervals between eruption rocks have a more potassic character, the Cucu (2.8–2.2 Ma; events. It is pointed out that the volcanic eruptions were often Pécskay et al., 1995), Pilişca (2.6–1.5 Ma; Pécskay et al., 1995)and separated by prolonged, i.e., N100 kyr quiescence periods. Recur- Murgul Mare volcanic rocks resemble the normal calc-alkaline series rence of volcanism even after such long episodes of dormancy of the CGH rather than the youngest CVDC, which comprises HKCAs has to be considered in assessing volcanic hazards, particularly volcanic products. Thus, another major change in the composition in seemingly inactive volcanic areas, where no Holocene erup- of the erupted magmas occurred at 1 Ma, when the first HKCAs dacite tions have occurred. The proposed term of “volcanoes with Po- and HKSs shoshonitic rocks erupted. The dominantly dacitic lava tentially Active Magma Storage” (Harangi et al., 2015a, 2015b) domes of the CVDC have a more pronounced Sr and Ba enrichment emphasizes the potential of volcanic rejuvenation in such volca- and depletion in the heavy REE (Seghedi et al., 1987; Szakács et al., nic regions. 1993; Mason et al., 1996; Vinkler et al., 2007), although they are compositionally heterogeneous. The lava domes of Malnaş and Bixad are distinct compared to the monotonous HKCAs dacites of Acknowledgements the CVDC. Although the change in the magma composition around 1–1.6 Ma has been previously inferred (e.g., Seghedi et al., 2011), This research was financed by the Hungarian National Research, our new combined geochronological and geochemical data provide Development and Innovation Fund (NKFIH) within No. K116528 more accurate constrains on it. and No. PD 121048 projects. The GATAN MiniCL facility belongs to The position of the Baba-Laposa lava dome has been debated: it has the KMP project nr. 4.2.1/B-10-2011-0002 supported by the been attributed to either Pilişca volcano (e.g., Szakács et al., 2015)orto European Union. Kata Molnár was subsidized by the Talented Ciomadul (e.g., Seghedi et al., 1987). Here, we propose that it represents Student Program of the Eötvös Loránd University (Budapest, the oldest member of the CVDC. The HKSs shoshonitic magmas were Hungary) and by the Campus Hungary Short Term Study Program previously considered to have erupted within or before the (Balassi Institute (B1/1R/14869), Hungary). I. Seghedi benefited by 1.6–1.0 Ma eruptive hiatus (e.g., Seghedi et al., 2011; Panaiotu et al., a grant of the Ministry of Education and Scientific Research, CNCS- 2012; Szakács et al., 2015). Our new data indicate that they extruded UEFISCFI, project number PN-II-IDPCE-2012-4-0137. The HIP facility much later, at 1000–900 ka. Remarkably, during a short period at Heidelberg University is operated under the auspices of the DFG (i.e., within a few 10s kyr), eruptions of three different magma types Scientific Instrumentation and Information Technology programme. (HKCAs and two distinct HKSs ones; Figs. 2 and 7B) occurred within a The invaluable help of Judit Dunklné Nagy (Göttingen) during the restricted area, a few km from one another. These volcanic products (U-Th)/He measurements is very much appreciated. The authors are distinct in chemical composition, but they have roughly similar min- thank the members of the MTA-ELTE Volcanological Research Group eral assemblage and textural features, showing evidence for open- (Budapest) for their helps during the field and laboratory works. K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 53

Appendix 1. Bulk rock composition of the studied southernmost Harghita samples

sample PD BL BH BAL NH MB-B MB-M NM KHM AB code in CSO-PD2 CSO-BL1 BH CSO-BH2 CSO-BAL CSO-BAL CSO-NH2 CSO-NH3 CSO-MB1 CSO-MB-M CSO-NM1 CSO-NM2 CSO-KHM1C CSO-KHM2A3 CSO-AB1 the text (cs)

SiO2 66.61 63.23 64.80 63.23 64.61 63.12 59.4 58.5 58.5 57.8 61.8 61.9 65.9 65.7 64.9

TiO2 0.56 0.45 0.53 0.54 0.53 0.52 0.7 0.8 0.9 0.9 0.7 0.7 0.4 0.3 0.4

Al2O3 17.64 17.26 17.30 17.81 17.77 17.50 18.1 18.4 15.5 16.3 18.3 17.7 16.1 16.2 16.9

Fe2O3 2.37 3.99 2.83 3.33 2.87 4.04 4.7 4.9 5.3 5.3 5.0 4.9 3.5 3.3 3.9 MnO 0.02 0.08 0.05 0.05 0.05 0.08 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 MgO 0.47 1.57 1.80 1.64 1.90 2.18 2.8 3.0 4.6 4.6 2.4 2.2 1.5 1.5 1.6 CaO 2.94 4.50 4.07 3.91 4.43 4.72 5.5 5.8 7.0 7.3 5.2 5.8 3.6 3.4 3.8

Na2O 3.87 4.06 4.48 4.26 4.51 4.51 4.1 4.1 3.6 3.9 4.0 4.2 4.1 4.1 4.4

K2O 2.70 3.22 3.27 3.27 3.29 3.15 2.6 2.6 4.0 3.5 2.4 2.4 3.4 3.2 3.2

P2O5 0.15 0.16 0.10 0.16 0.14 0.17 0.2 0.2 0.5 0.3 0.2 0.2 0.1 0.1 0.2 Cr 14 14 110 21 131 103 21 145 164 34 27 27 27 34 Ni b20 b20 b20 b20 b20 b20 b20 b20 45 32 b20 b20 b20 b20 b20 Sc87 888 8 1213131614135 5 6 Ba 1284 1429 1472 1509 1493 1534 1164 1215 2707 1454 948 847 1283 1306 1506 Be3223322 b12 3 2 1 2 3 3 Co 4.4 5.7 5.9 6.1 6.6 8.0 8.7 9.2 15.5 21.9 12.4 10.7 5.9 6 7 Cs 12.4 4.5 3.3 3.7 3.8 2.1 2.8 3.0 1.6 0.8 2.1 1.9 4.2 4 13 Ga 16.0 18.3 19.3 18.1 18.5 19.2 18.2 19.4 18.1 18.5 17.6 17.3 17.3 17 18 Hf 3.1 3.6 3.9 3.7 3.6 3.4 3.3 3.5 6.6 4.8 3.7 3.5 4.0 4 4 Nb 17.1 14.8 15.5 16.5 13.9 14.4 15.5 15.4 17.6 18.9 19.7 18.0 13.4 13 15 Rb 87.2 83.4 82.4 85.6 81.7 71.4 58.2 58.3 69.0 58.2 73.9 72.8 95.6 96 86 Sn 1 1 2 1 2 b12211 1 b12 b11 Sr 844.8 1209.3 1316.4 1264.5 1281.5 1354.2 1167.4 1248.6 2509.5 1713.7 696.8 732.1 1221.1 1213 1416 Ta 1.2 1.0 0.8 1.1 0.9 0.8 1.0 1.0 1.1 1.1 1.5 1.3 1.0 1 1 Th 12.6 13.6 11.9 12.2 11.4 11.3 9.0 9.7 14.8 10.0 10.8 9.5 14.9 14 14 U 3.9 4.1 3.1 3.3 3.2 3.2 2.8 2.7 3.9 2.7 3.0 3.0 4.6 5 4 V 76 61 61 75 64 64 101 112 108 117 79 94 45 46 46 W 1.2 1.1 0.9 0.7 0.9 0.5 1.4 0.7 1.1 2.8 1.2 3.2 1.5 1 2 Zr 126.3 134.9 145.9 148.9 141.9 137.1 127.5 133.5 255.0 197.9 146.8 139.1 148.1 145 158 Y 27.8 12.3 10.1 11.3 11.0 11.8 12.0 13.0 16.2 14.8 14.0 15.8 9.1 10 10 La 47.9 35.1 36.8 41.4 39.0 38.1 30.2 31.3 108.0 49.9 31.9 27.8 33.9 33 43 Ce 86.9 63.2 59.8 65.8 67.0 66.2 52.7 59.0 199.8 96.2 51.9 44.9 57.5 56 71 Pr 8.94 6.15 5.84 6.24 6.81 6.66 5.51 5.72 22.21 10.68 5.16 4.83 5.67 5 6 Nd 34.70 22.50 19.20 21.80 23.50 22.60 21.00 21.70 82.90 39.60 18.30 17.70 20.40 19 21 Sm 6.84 3.32 3.02 3.43 3.42 3.40 3.25 3.51 10.89 5.76 2.82 3.06 2.94 3 3 Eu 2.19 0.96 0.94 1.04 1.03 1.03 1.08 1.08 2.73 1.53 0.86 0.97 0.91 1 1 Gd 6.77 2.74 2.28 2.59 2.65 2.86 2.93 3.05 6.85 4.06 2.73 2.98 2.40 2 2 Tb 1.02 0.38 0.31 0.35 0.37 0.35 0.39 0.46 0.75 0.49 0.42 0.46 0.31 0 0 Dy 5.84 2.13 1.61 1.87 2.09 1.97 2.18 2.18 3.74 2.61 2.42 2.88 1.79 2 2 Ho 1.24 0.44 0.33 0.38 0.38 0.41 0.43 0.47 0.57 0.49 0.47 0.56 0.34 0 0 Er 3.28 1.17 0.98 1.18 1.26 1.17 1.25 1.26 1.49 1.39 1.50 1.69 1.00 1 1 Tm 0.52 0.21 0.15 0.20 0.16 0.18 0.19 0.18 0.25 0.20 0.22 0.25 0.16 0 0 Yb 3.34 1.44 1.13 1.26 1.08 1.18 1.15 1.23 1.47 1.36 1.41 1.65 1.12 1 1 Lu 0.49 0.22 0.17 0.21 0.19 0.20 0.20 0.22 0.22 0.19 0.25 0.27 0.16 0 0 LOI 2.7 1.5 0.8 1.8 -0.1 0.3 1.9 1.8 0.0 0.8 1.5 1.1 1.5 2.0 0.8

Appendix 2. “Secular equilibrium-assumed”-, combined U-Pb/U-Th- and “full disequilibrium-assumed” (U-Th)/He ages of the studied samples with their 1σ uncertainties, weighted means and D230 values used for the calculation

Secular eq. assumed (U-Th)/He 1σ D230 Combined U-Pb / U-Th and (U-Th)/He +1s −1s Full diseq. assumed (U-Th)/He 1σ

KHM: 163 ± 11 (0.321) CSO-KHM-1 z1 145.1 12.9 0.26 178.9 16.7 20.2 186.2 16.5 CSO-KHM-1 z2 127.8 9.5 0.20 154.3 17.6 13.1 172.2 12.9 CSO-KHM-1 z3 132.1 9.8 0.26 158.6 15.4 14.8 172.2 12.9 CSO-KHM-1 z5 144.4 9.9 0.25 177.4 14.2 15.2 186.2 12.8 CSO-KHM-1 z6 160.7 10.0 0.31 194.3 12.0 13.2 199.4 12.4 CSO-KHM1 z10 131.0 10.7 0.22 160.1 17.6 16.5 174.7 14.3 CSO-KHM1 z11 118.1 10.0 0.21 141.3 15.3 14.7 160.6 13.6 CSO-KHM1 z12 140.3 13.4 0.21 176.4 18.7 22.0 185.2 17.8 in_CSO-KHM-1 z12 127.3 11.5 0.23 155.5 16.6 19.1 169.6 15.3 in_CSO-KHM-1 z13 119.6 11.8 0.20 144.6 18.2 17.9 163.6 16.2 in_CSO-KHM-1 z5 123.4 9.1 0.24 147.7 16.3 16.4 164.3 12.2 in_CSO-KHM-1 z6 131.6 10.1 0.38 151.0 15.3 16.6 162.3 12.5 Weighted means: 133.1 3.5 174.7 4.6

(continued on next page) 54 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

(continued)

Secular eq. assumed (U-Th)/He 1σ D230 Combined U-Pb / U-Th and (U-Th)/He +1s −1s Full diseq. assumed (U-Th)/He 1σ

AB: 344 ± 33 ka (0.782) CSO-AB z1 333.0 27.2 0.20 351.6 31.1 34.1 387.7 31.7 CSO-AB z2 338.0 24.4 0.23 355.3 29.3 29.9 389.7 28.2 CSO-AB z3 315.6 22.3 0.21 330.4 24.8 27.1 369.0 26.1 Weighted means: 327.7 14.1 382.1 6.6

BAL: 583 ± 30 ka (0.097) MK-2 z1 575.0 31.3 0.15 578.8 33.9 32.4 636.4 34.8 MK-2 z2 545.5 36.5 0.18 545.6 40.7 34.5 603.6 40.5 MK-2 z3 624.3 43.2 0.21 626.7 50.3 41.9 679.4 47.1 MK-2 z4 637.7 43.1 0.15 647.9 44.7 48.9 699.2 47.3 MK-2 z5 525.3 35.2 0.15 524.5 39.9 32.4 586.6 39.4 MK-2 z6 596.1 34.1 0.18 598.2 39.6 33.1 654.3 37.5 CSO-BAL(cs) z1 642.9 48.5 0.17 646.1 58.0 47.6 702.2 53.1 CSO-BAL(cs) z2 509.0 30.1 0.19 508.8 33.5 28.6 566.0 33.5 CSO-BAL(cs) z3 647.3 40.1 0.14 651.6 50.1 39.3 709.9 44.1 CSO-BAL(cs) z4 594.0 40.8 0.20 595.4 46.3 39.5 650.1 44.7 CSO-BAL(cs) z5 586.2 33.3 0.22 587.1 38.7 31.4 640.3 36.5 Weighted means: 579.7 18.4 648.0 14.5

BH: 642 ± 44 ka (0.219) CSO-BH z1 723.3 61.7 0.30 741.7 73.5 73.2 769.8 65.7 CSO-BH z2 654.4 43.6 0.27 659.7 50.9 44.3 703.7 47.0 CSO-BH z3 576.8 34.0 0.17 579.9 38.2 33.6 636.0 37.6 CSO-BH z4 714.4 57.3 0.26 732.8 69.9 68.6 764.7 61.5 CSO-BH z6 689.4 56.5 0.16 699.6 74.9 60.4 749.8 61.5 CSO-BH(sz) z1 585.4 41.7 0.20 586.5 47.7 40.2 641.5 45.8 CSO-BH(sz) z2 657.7 51.4 0.17 663.0 63.9 52.2 717.0 56.2 CSO-BH(sz) z3 675.6 51.2 0.09 686.3 67.9 54.8 743.9 56.4 Weighted means: 640.5 24.2 715.8 18.6

NH: 842 ± 53 ka (0.927) CSO-NH5 z1 878.5 71.0 0.29 885.4 99.1 74.8 926.0 75.0 CSO-NH5 z2 821.1 52.5 0.19 828.4 58.7 55.2 878.4 56.3 CSO-NH5 z3 816.3 65.0 0.20 820.5 71.2 66.0 872.5 69.6 CSO-NH5 z4 816.1 56.0 0.57 815.1 63.0 53.3 841.3 57.9 CSO-NH5 z5 873.5 48.6 0.17 888.1 68.1 54.9 932.9 52.1 CSO-NH5 z6 829.5 60.4 0.19 836.4 69.2 63.5 886.8 64.7 Weighted means: 839.4 23.5 889.6 14.1

MB-B: 907 ± 66 ka (0.006) CSO-MB-B z1 783.4 45.9 0.17 788.9 42.2 51.6 843.1 49.6 CSO-MB-B z2 841.9 45.3 0.15 843.3 47.2 46.5 903.5 48.8 CSO-MB-B z3 943.6 53.8 0.22 942.9 60.3 51.2 997.5 57.0 CSO-MB-B z4 886.8 54.9 0.14 896.2 49.9 64.0 949.0 58.9 CSO-MB-B z5 1062.1 83.6 0.21 1067.4 98.4 87.4 1117.1 88.0 CSO-MB-B z6 966.7 77.7 0.17 964.7 87.3 76.1 1026.3 82.6 MB z8 1026.8 64.9 0.17 1030.1 72.7 65.9 1086.0 68.9 MB z9 1000.8 63.4 0.17 1003.6 70.3 63.6 1060.5 67.3 Weighted means: 906.8 33.8 997.9 35.6

BL: 942 ± 65 ka (0.044) CSO-BL-1 z2 939.5 58.8 0.22 942.8 66.2 58.5 993.5 62.4 CSO-BL-1 z3 1041.9 64.6 0.52 1047.8 83.9 63.4 1070.8 66.6 CSO-BL-1 z4 983.4 71.1 0.20 989.8 94.3 74.6 1040.1 75.3 CSO-BL-1 z5 1056.5 77.1 0.21 1101.2 79.5 109.5 1111.9 81.3 CSO-BL-1 z6 978.9 70.0 0.23 986.7 85.4 75.7 1032.6 74.0 CSO-BL-1 zR2 943.4 67.0 0.21 949.0 73.4 71.0 998.6 71.0 CSO-BL-1 zR4 782.1 50.5 0.34 779.7 55.2 47.3 825.0 53.3 CSO-BL-1 zR5 1046.2 85.7 0.18 1107.4 73.7 140.4 1104.3 90.6 CSO-BL-1 zR6 924.3 82.1 0.26 925.5 87.3 82.6 974.5 86.7 CSO-BL-1 zR7 857.5 57.0 0.23 858.1 63.2 56.9 910.4 60.6 CSO-BL-1 zR8 1047.4 94.6 0.31 1093.3 89.5 139.0 1093.3 98.8 Weighted means: 939.1 27.6 1014.1 27.7

MB-M: 964 ± 46 ka (0.871) CSO-MB-M z4 1049.2 80.0 0.24 1047.1 87.0 77.5 1101.5 84.3 CSO-MB-M z5 966.9 56.6 0.21 963.0 63.8 52.5 1022.2 60.3 CSO-MB-M z6 921.4 63.2 0.22 917.1 69.9 58.5 975.6 67.2 CSO-MB-M z7 994.8 69.5 0.17 992.1 76.3 65.7 1054.2 74.0 CSO-MB-M z8 933.5 54.2 0.21 940.1 50.7 60.9 988.7 57.8 CSO-MB-M z9 937.6 64.3 0.19 944.7 60.2 71.7 994.9 68.5 CSO-MB-M z10 944.3 66.0 0.20 938.6 74.2 60.3 1000.6 70.3 CSO-MB-M z11 1033.9 73.1 0.21 1031.8 80.3 70.6 1089.2 77.3 Weighted means: 964.9 22.8 1028.4 16.9

PD: 1640 ± 37 ka CSO-PD2 z4 1597.1 117.0 1651.5 121.2 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56 55

(continued)

Secular eq. assumed (U-Th)/He 1σ D230 Combined U-Pb / U-Th and (U-Th)/He +1s −1s Full diseq. assumed (U-Th)/He 1σ

CSO-PD2 z5 1601.5 121.3 1658.9 125.9 CSO-PD2 z6 1649.5 135.5 1698.1 139.7 CSO-PD2 z7 1696.7 116.4 1756.3 120.7 CSO-PD2 z8 1664.5 146.9 1715.9 151.7 Weighted means: 1640.3 37.2 1696.1 21.5

NM: 1865 ± 87 ka CSO-NM1 z1 1856.9 129.7 1924.3 135.1 CSO-NM1 z2 1947.8 125.9 2012.9 130.8 CSO-NM1 z3 1886.6 137.7 1952.9 143.2 CSO-NM1 z4 1761.5 99.2 1827.7 103.7 CSO-NM1 z5 2068.1 160.2 2133.3 165.9 CSO-NM1 z6 1807.9 135.5 1874.1 141.1 Weighted means: 1865.5 87.2 1954.2 48.5

Concordant eruption ages with 2σ uncertainty and goodness of ﬁt parameter (in brackets) are recorded next to the sample names.

References Harangi, S., 2001. Neogene to Quaternary volcanism of the Carpathian–Pannonian Re- gion — a review. Acta Geol. Hung. 44, 223–258. Braun, J., van der Beek, P., Batt, G., 2012. Quantitative Thermochronology: Numerical Harangi, S., 2007. A Kárpát-Pannon térség legutolsó vulkáni kitörései - lesz-e még Methods for the Interpretation of Thermochronological Data. Cambridge University folytatás? (The last volcanic eruptions in the Carpathian-Pannonian Region - to be Press (272 pp). continued?). Földrajzi Közlemények 131 (4), 271–288. Casta, L., 1980. Les formation quaternaires de la depression de Brasov (Roumaine). Harangi, S., Lenkey, L., 2007. Genesis of the Neogene to Quaternary volcanism in the Université d'Aix Marseille II (PhD Thesis Thesis). Carpathian-Pannonian region: role of subduction, extension, and mantle plume. Chalot-Prat, F., Gîrbacea, R., 2000. Partial delamination of continental mantle lithosphere, Geol. Soc. Am. Spec. Pap. 418:67–92. https://doi.org/10.1130/2007.2418(04). uplift-related crust-mantle decoupling, volcanism and basin formation: a new model Harangi, S., Downes, H., Thirlwall, M., Gméling, K., 2007. Geochemistry, petrogenesis and for the Pliocene-Quaternary evolution of the southern East-Carpathians, Romania. geodynamic relationships of Miocene calc-alkaline volcanic rocks in the Western Tectonophysics 327, 83–107. Carpathian Arc, Eastern Central Europe. J. Petrol. 48 (12):2261–2287. https://doi. Ciulavu, D., Dinu, C., Szakács, A., Dordea, D., 2000. Neogene kinematics of the Transylva- org/10.1093/petrology/egm059. nian basin (Romania). AAPG Bull. 84 (10), 1589–1615. Harangi, S., Molnár, M., Vinkler, A.P., Kiss, B., Jull, A.J.T., Leonard, A.G., 2010. Radiocarbon Cloetingh, S.A.P.L., Burov, E., Matenco, L., Toussaint, G., Bertotti, G., Andriessen, P.A.M., dating of the last volcanic eruptions of Ciomadul volcano, Southeast Carpathians, Wortel, M.J.R., Spakman, W., 2004. Thermo-mechanical controls on the mode of con- Eastern-Central Europe. Radiocarbon 52 (2–3), 1498–1507. tinental collision in the SE Carpathians (Romania). Earth Planet. Sci. Lett. 218 (1–2): Harangi, S., Sági, T., Seghedi, I., Ntaflos, T., 2013. Origin of basaltic magmas of Perşani vol- 57–76. https://doi.org/10.1016/S0012-821X(03)00645-9. canic field, Romania: a combined whole rock and mineral scale investigation. Lithos Clynne, M.A., Muffler, L.J.P., 2010. Geologic map of Lassen Volcanic National Park and Vi- 180–181 (0):43–57. https://doi.org/10.1016/j.lithos.2013.08.025. cinity, California. U.S. Geological Survey Scientific Investigations Map 2899 (scale 1: Harangi, S., Lukács, R., Schmitt, A.K., Dunkl, I., Molnár, K., Kiss, B., Seghedi, I., Novothny, Á., 50,000). Molnár, M., 2015a. Constraints on the timing of Quaternary volcanism and duration Connor,C.B.,McBirney,A.R.,Furlan,C.,2006.What Is the Probability of Explosive Eruption at of magma residence at Ciomadul volcano, east–central Europe, from combined U– a Long-dormant Volcano? School of Geosciences Faculty and Staff Publication, p. 1041 Th/He and U–Th zircon geochronology. J. Volcanol. Geotherm. Res. 301:66–80. Csontos, L., Nagymarosy, A., Horváth, F., Kovác, M., 1992. Tertiary evolution of the Intra- https://doi.org/10.1016/j.jvolgeores.2015.05.002. Carpathian area: a model. Tectonophysics 208 (1–3), 221–241. Harangi, S., Novák, A., Kiss, B., Seghedi, I., Lukács, R., Szarka, L., Wesztergom, V., Metwaly, Danišík, M., Shane, P., Schmitt, A.K., Hogg, A., Santos, G.M., Storm, S., Evans, N.J., Keith M., Gribovszki, K., 2015b. Combined magnetotelluric and petrologic constrains for the Fifield, L., Lindsay, J.M., 2012. Re-anchoring the late Pleistocene tephrochronology nature of the magma storage system beneath the Late Pleistocene Ciomadul volcano of New Zealand based on concordant radiocarbon ages and combined 238U/230Th (SE Carpathians). J. Volcanol. Geotherm. Res. 290:82–96. https://doi.org/10.1016/j. disequilibrium and (U–Th)/He zircon ages. Earth Planet. Sci. Lett. 349–350: jvolgeores.2014.12.006. 240–250. https://doi.org/10.1016/j.epsl.2012.06.041. Hildenbrand, A., Marques, F.O., Costa, A.C.G., Sibrant, A.L.R., Silva, P.F., Henry, B., Miranda, Danišík,M.,Schmitt,A.K.,Stockli,D.F.,Lovera,O.M.,Dunkl,I.,Evans,N.J.,2017.Ap- J.M., Madureira, P., 2012. Reconstructing the architectural evolution of volcanic plication of combined U-Th-disequilibrium/U-Pb and (U-Th)/He zircon dating islands from combined K/Ar, morphologic, tectonic, and magnetic data: the Faial Is- to tephrochronology. Quat. Geochronol. 40:23–32. https://doi.org/10.1016/j. land example (Azores). J. Volcanol. Geotherm. Res. 241–242:39–48. https://doi.org/ quageo.2016.07.005. 10.1016/j.jvolgeores.2012.06.019. Downes, H., Seghedi, I., Szakács, A., Dobosi, G., James, D.E., Vaselli, O., Rigby, I.J., Ingram, Hora, J.M., Singer, B.S., Wörner, G., 2007. Volcano evolution and eruptive flux on the thick G.A., Rex, D., Pécskay, Z., 1995. Petrology and geochemistry of late Tertiary/Quater- crust of the Andean Central Volcanic Zone: 40Ar/39Ar constraints from Volcán nary mafic alkaline volcanism in Romania. Lithos 35, 65–81. Parinacota, Chile. Geol. Soc. Am. Bull. 119 (3–4):343–362. https://doi.org/10.1130/ Farley, K.A., 2002. (U-Th)/He dating: techniques, calibrations, and applications. Rev. Min- B25954.1. eral. Geochem. 47 (1):819–844. https://doi.org/10.2138/rmg.2002.47.18. Horváth, F., Bada, G., Szafián, P., Tari, G., Ádám, A., Cloetingh, S., 2006. Formation and de- Farley, K.A., Wolf, R.A., Silver, L.T., 1996. The effects of long alpha-stopping distances on formation of the Pannonian Basin: constraints from observational data. Geol. Soc. (U-Th)/He ages. Geochim. Cosmochim. Acta 60 (21):4223–4229. https://doi.org/ Lond. Mem. 32 (1):191–206. https://doi.org/10.1144/gsl.mem.2006.032.01.11. 10.1016/S0016-7037(96)00193-7. Hourigan, J.K., Reiners, P.W., Brandon, M.T., 2005. U-Th zonation-dependent alpha- Farley, K.A., Kohn, B.P., Pillans, B., 2002. The effects of secular disequilibrium on (U–Th)/ ejection in (U-Th)/He chronometry. Geochim. Cosmochim. Acta 69 (13): He systematics and dating of Quaternary volcanic zircon and apatite. Earth Planet. 3349–3365. https://doi.org/10.1016/j.gca.2005.01.024. Sci. Lett. 201 (1):117–125. https://doi.org/10.1016/S0012-821X(02)00659-3. Ianovici, V. and Rădulescu, D. 1966. Geological map 1:200000 L-35-XIV, sheet 20 Odorhei. Fielitz, W., Seghedi, I., 2005. Late Miocene–Quaternary volcanism, tectonics and drainage Geological Institute of Romania, Bucharest. system evolution in the East Carpathians, Romania. Tectonophysics 410 (1–4): Ismail-Zadeh, A., Matenco, L., Radulian, M., Cloetingh, S., Panza, G., 2012. Geodynamics 111–136. https://doi.org/10.1016/j.tecto.2004.10.018. and intermediate-depth seismicity in Vrancea (the south-eastern Carpathians): cur- Fillerup, M.A., Knapp, J.H., Knapp, C.C., Raileanu, V., 2010. Mantle earthquakes in the ab- rent state-of-the art. Tectonophysics 530–531:50–79. https://doi.org/10.1016/j. sence of subduction? Continental delamination in the Romanian Carpathians. Litho- tecto.2012.01.016. sphere 2 (5):333–340. https://doi.org/10.1130/L102.1. Karátson, D., Telbisz, T., Harangi, S., Magyari, E., Dunkl, I., Kiss, B., Jánosi, C., Veres, D., Braun, Gamble, J.A., Price, R.C., Smith, I.E.M., McIntosh, W.C., Dunbar, N.W., 2003. 40Ar/39Ar geo- M., Fodor, E., Biró, T., Kósik, S., von Eynatten, H., Lin, D., 2013. Morphometrical and geochronology of magmatic activity, magma flux and hazards at Ruapehu volcano, Taupo chronological constraints on the youngest eruptive activity in East-Central Europe at the Volcanic Zone, New Zealand. J. Volcanol. Geotherm. Res. 120 (3):271–287. https:// Ciomadul (Csomád) lava dome complex, East Carpathians. J. Volcanol. Geotherm. Res. doi.org/10.1016/S0377-0273(02)00407-9. 255:43–56. https://doi.org/10.1016/j.jvolgeores.2013.01.013. Ghalamghash, J., Mousavi, S.Z., Hassanzadeh, J., Schmitt, A.K., 2016. Geology, zircon geo- Karátson, D., Wulf, S., Veres, D., Magyari, E.K., Gertisser, R., Timar-Gabor, A., Novothny, Á., chronology, and petrogenesis of Sabalan volcano (northwestern Iran). J. Volcanol. Telbisz, T., Szalai, Z., Anechitei-Deacu, V., Appelt, O., Bormann, M., Jánosi, C., Hubay, K., Geotherm. Res. 327:192–207. https://doi.org/10.1016/j.jvolgeores.2016.05.001. Schäbitz, F., 2016. The latest explosive eruptions of Ciomadul (Csomád) volcano, East Gîrbacea, R., Frisch, W., 1998. Slab in the wrong place: lower lithospheric mantle delam- Carpathians — a tephrostratigraphic approach for the 51–29 ka BP time interval. ination in the last stage of Eastern Carpathians subduction retreat. Geology 26, J. Volcanol. Geotherm. Res. 319:29–51. https://doi.org/10.1016/j.jvolgeores.2016.03.005. 611–614. Kis, B.-M., Ionescu, A., Cardellini, C., Harangi, S., Baciu, C., Caracausi, A., Viveiros, F., 2017. Girbacea, R., Frisch, W., Linzer, H.-G., 1998. Post-orogenic uplift-induced extension: a ki- Quantification of carbon dioxide emissions of Ciomadul, the youngest volcano of nematic model for the Pliocene to recent tectonic evolution of the Eastern the Carpathian-Pannonian Region (Eastern-Central Europe, Romania). J. Volcanol. Carpathians (Romania). Geol. Carpath. 49 (5), 315–327. Geotherm. Res. https://doi.org/10.1016/j.jvolgeores.2017.05.025. 56 K. Molnár et al. / Journal of Volcanology and Geothermal Research 354 (2018) 39–56

Kiss, B., Harangi, S., Ntaflos, T., Mason, P.R.D., Pál-Molnár, E., 2014. Amphibole perspective Reiners, P.W., 2005. Zircon (U-Th)/He thermochronometry. Rev. Mineral. Geochem. 58 to unravel pre-eruptive processes and conditions in volcanic plumbing systems be- (1):151–179. https://doi.org/10.2138/rmg.2005.58.6. neath intermediate arc volcanoes: a case study from Ciomadul volcano (SE Royden, L., Horváth, F., Rumpler, J., 1983. Evolution of the Pannonian Basin system: 1. Tec- Carpathians). Contrib. Mineral. Petrol. 167 (3):1–27. https://doi.org/10.1007/ tonics. Tectonics 2 (1):63–90. https://doi.org/10.1029/TC002i001p00063. s00410-014-0986-6. Schmitt, A.K., 2011. Uranium series accessory crystal dating of magmatic processes. Annu. Klemetti, E.W., Grunder, A.L., 2008. Volcanic evolution of Volcán Aucanquilcha: a long- Rev. Earth Planet. Sci. 39 (1):321–349. https://doi.org/10.1146/annurev-earth- lived dacite volcano in the Central Andes of northern Chile. Bull. Volcanol. 70 (5): 040610-133330. 633–650. https://doi.org/10.1007/s00445-007-0158-x. Schmitt, A.K., Stockli, D.F., Hausback, B.P., 2006. Eruption and magma crystallization ages Konečný, V., Kovác, M., Lexa, J., Šefara, J., 2002. Neogene evolution of the Carpatho- of Las Tres Vírgenes (Baja California) constrained by combined 230Th/238U and (U– Pannonian region: an interplay of subduction and back-arc diapiric uprise in the Th)/He dating of zircon. J. Volcanol. Geotherm. Res. 158 (3–4):281–295. https://doi. mantle. European Geophysical Union Stephan Mueller Special Publication Series org/10.1016/j.jvolgeores.2006.07.005. vol. 1, pp. 105–123. Schmitt, A.K., Stockli, D.F., Niedermann, S., Lovera, O.M., Hausback, B.P., 2010. Eruption Koulakov, I., Zaharia, B., Enescu, B., Radulian, M., Popa, M., Parolai, S., Zschau, J., 2010. De- ages of Las Tres Vírgenes volcano (Baja California): a tale of two helium isotopes. lamination or slab detachment beneath Vrancea? New arguments from local earth- Quat. Geochronol. 5 (5):503–511. https://doi.org/10.1016/j.quageo.2010.02.004. quake tomography. Geochem. Geophys. Geosyst. 11 (3). https://doi.org/10.1029/ Seghedi, I., Downes, H., 2011. Geochemistry and tectonic development of Cenozoic 2009gc002811. magmatism in the Carpathian–Pannonian region. Gondwana Res. 20 (4):655–672. Lindsay, J.M., Trumbull, R.B., Schmitt, A.K., Stockli, D.F., Shane, P.A., Howe, T.M., 2013. Vol- https://doi.org/10.1016/j.gr.2011.06.009. canic stratigraphy and geochemistry of the Soufrière Volcanic Centre, Saint Lucia with Seghedi, I., Szakács, A., Udrescu, C., Stoian, M., Grabari, G., 1987. Trace element geochem- implications for volcanic hazards. J. Volcanol. Geotherm. Res. 258:126–142. https:// istry of the South Hargitha volcanics (East Carpathians): calc-alkaline and shoshonitic doi.org/10.1016/j.jvolgeores.2013.04.011. association. Dari de Seama ale Sedintelor Institutul de Geologie si Geofizica 72-73, Lukács, R., Guillong, M., Schmitt, A.K., Molnár, K., Harangi, Sz., Bachmann, O., 2018. 381–397. LA-ICP-MS and SIMS U-Pb and U-Th zircon geochronology of Late Pleistocene Seghedi, I., Balintoni, I., Szakács, A., 1998. Interplay of tectonics and neogene post- lava domes of the Ciomadul Volcanic Dome Complex (Eastern Carpathians). collisional magmatism in the intracarpathian region. Lithos 45 (1–4):483–497. Data Brief (submitted). https://doi.org/10.1016/S0024-4937(98)00046-2. Martin, M., Wenzel, F., CALIXTO, W.G., 2006. High-resolution teleseismic body wave to- Seghedi, I., Downes, H., Szakács, A., Mason, P.R.D., Thirlwall, M.F., Rosu, E., Pécskay, Z., mography beneath SE-Romania – II. Imaging of a slab detachment scenario. Geophys. Marton, E., Panaiotu, C., 2004. Neogene-Quaternary magmatism and geodínamics in J. Int. 164 (3):579–595. https://doi.org/10.1111/j.1365-246X.2006.02884.x. the Carpathian-Pannonian region: a synthesis. Lithos 72, 117–146. Mason, P.R.D., Downes, H., Seghedi, I., Szakács, A., Thirlwall, M.F., 1995. Low-pressure evo- Seghedi, I., Maţenco, L., Downes, H., Mason, P.R.D., Szakács, A., Pécskay, Z., 2011. Tectonic lution of magmas from the Calimani, Gurghiu and Harghita Mountains, East significance of changes in post-subduction Pliocene–Quaternary magmatism in the Carpathians. Acta Vulcanol. 7 (2), 43–52. south east part of the Carpathian–Pannonian Region. Tectonophysics 502 (1–2): Mason, P.R.D., Downes, H., Thirlwall, M., Seghedi, I., Szakács, A., Lowry, D., Mattey, D., 146–157. https://doi.org/10.1016/j.tecto.2009.12.003. 1996. Crustal assimilation as a major petrogenetic process in east Carpathian Neo- Seghedi, I., Popa, R.-G., Panaiotu, C.G., Szakács, A., Pécskay, Z., 2016. Short-lived eruptive gene to Quaternary continental margin arc magmas. J. Petrol. 37, 927–959. episodes during the construction of a Na-alkalic basaltic field (Perşani Mountains, Mason, P.R.D., Seghedi, I., Szakács, A., Downes, H., 1998. Magmatic constraints on SE Transylvania, Romania). Bull. Volcanol. 78 (10):69. https://doi.org/10.1007/ geodynamic models of subduction int he East Carpathians, Romania. Tectonophysics s00445-016-1063-y. 297, 157–176. Siebert, L., Simkin, T., Kimberly, P., 2011. Volcanoes of the World. University of California Moriya, I., Okuno, M., Nakamura, E., Szakács, A., Seghedi, I., 1995. Last eruption and its 14C Press (568 pp). age of Ciomadul volcano, Romania. Summaries of Researches Using AMS at Nagoya Singer, B.S., 2014. A Quaternary geomagnetic instability time scale. Quat. Geochronol. 21: University vol. 6, pp. 82–91. 29–52. https://doi.org/10.1016/j.quageo.2013.10.003. Moriya, I., Okuno, M., Nakamura, T., Ono, K.A.S., Seghedi, I., 1996. Radiocarbon ages of Sperner, B., Lorenz, F., Bonjer, K., Hettel, S., Muller, B., Wenzel, F., 2001. Slab break-off - charcoal fragments from the pumice flow deposits of the last eruption of Ciomadul aprupt cut or gradual detachment? New insights from the Vrancea Region (SE volcano, Romania. Summaries of Researches Using AMS at Nagoya University vol. Carpathians, Romania). Terra Nova 13, 172–179. VII(3), pp. 252–255. Sun, S.-s., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: Mucek, A.E., Danišík, M., de Silva, S.L., Schmitt, A.K., Pratomo, I., Coble, M.A., 2017. Post- implications for mantle composition and processes. Geol. Soc. Lond., Spec. Publ. 42 supereruption recovery at Toba Caldera. Nat. Commun. 8, 15248. https://doi.org/ (1):313–345. https://doi.org/10.1144/gsl.sp.1989.042.01.19. 10.1038/ncomms15248. Szabó, C., Harangi, S., Csontos, L., 1992. Review of neogene and quaternary volcanism of Nakamura, N., 1974. Determination of REE, Ba, Fe, Mg, Na and K in carbonaceous and or- the Carpathian-Pannonian region: a review. Tectonophysics 208, 243–256. dinary chondrites. Geochim. Cosmochim. Acta 38 (5):757–775. https://doi.org/ Szakács, A., Seghedi, I., 1986. Chemical diagnosis of the volcanics from the southeastern- 10.1016/0016-7037(74)90149-5. most part of the Harghita Mountains — proposal for a new nomenclature. Rev. Roum. Oncescu, M.C., Burlacu, V., Anghel, M., Smalbergher, V., 1984. Three-dimensional P-wave Géol. 30, 41–48. velocity image under the Carpathian arc. Tectonophysics 106 (3):305–319. https:// Szakács, A., Seghedi, I., 2013. The relevance of volcanic hazard in Romania: is there any? En- doi.org/10.1016/0040-1951(84)90182-3. viron. Eng. Manag. J. 12, 125–135. Panaiotu, C.G., Vişan, M., Ţugui, A., Seghedi, I., Panaiotu, A.G., 2012. Palaeomagnetism of Szakács, A., Seghedi, I., Pécskay, Z., 1993. Pecularities of South Hargitha Mts. as the termi- the South Harghita volcanic rocks of the East Carpathians: implications for tectonic nal segment of the Carpathian Neogene to Quaternary volcanic chain. Revue rotations and palaeosecular variation in the past 5 Ma. Geophys. J. Int. 189 (1): Roumaine de Géologie Géophysique et Géographie, Géologie 37, 21–37. 369–382. https://doi.org/10.1111/j.1365-246X.2012.05394.x. Szakács, A., Ioane, D., Seghedi, I., Rogobete, M., Pécskay, Z., 1997. Rates of migration of vol- Peccerillo, A., Taylor, S.R., 1976. Geochemistry of eocene calc-alkaline volcanic rocks from canic activity and magma output along the Calimani-Gurghiu-Harghita volcanic the Kastamonu area, Northern Turkey. Contrib. Mineral. Petrol. 58 (1):63–81. https:// range. East Carpathians, PANCARDI'97. Prz. Geol. 45 (10/2), 1106. doi.org/10.1007/bf00384745. Szakács, A., Seghedi, I., Pécskay, Z., 2002. The most recent volcanism in the Carpathian- Pécskay, Z., Szakács, A., Seghedi, I., Karátson, D., 1992. Contributions to the geochronology Pannonian Region. Is there any volcanic hazard? Geologica Carpathica Special Issue, of Mt. Cucu volcano and the South Harghita (East Carpathians, Romania). Földtani Proceedings of the XVIIth Congress of Carpathian-Balkan Geological Association vol. Közlöny 122 (2–4), 265–286. 53, pp. 193–194 Pécskay, Z., Lexa, J.A.S., Balogh, K., Seghedi, I., Konecny, V., Kovács, M., Márton, E., Kaliciak, Szakács, A., Seghedi, I., Pécskay, Z., Mirea, V., 2015. Eruptive history of a low-frequency M., Széky-Fux, V., Póka, T., Gyarmati, P., Edelstein, O., Rosu, E., Zec, B., 1995. Space and and low-output rate Pleistocene volcano, Ciomadul, South Harghita Mts., Romania. time distribution of Neogene-Quaternary volcanism in the Carpatho-Pannonian Re- Bull. Volcanol. 77 (2):1–19. https://doi.org/10.1007/s00445-014-0894-7. gion. In: Downes, H., Vaselli, O. (Eds.), Neogene and Related Magmatism in the Upton, G., Cook, I., 1996. Understanding Statistics. Oxford University Press. Carpatho-Pannonian Region. Acta Vulcanologica, pp. 15–28. Vaselli, O., Minissale, A., Tassi, F., Magro, G., Seghedi, I., Ioane, D., Szakács, A., 2002. Ageo- Pécskay, Z., Lexa, J., Szakács, A., Seghedi, I., Balogh, K., Konecný, V., Zelenka, T., Kovacs, M., chemical traverse across the Eastern Carpathians (Romania): constraints on the ori- Póka, T., Fülöp, A., Márton, E., Panaiotu, C., Cvetkovic, V., 2006. Geochronology of Neo- gin and evolution of the mineral water and gas discharges. Chem. Geol. 182, 637–654. gene magmatism in the Carpathian arc and intra-Carpathian area. Geol. Carpath. 57 Vinkler, A.P., Harangi, S., Ntaflos, T., Szakács, A., 2007. A Csomád vulkán (Keleti-Kárpátok) (6), 511–530. horzsaköveinek kőzettani és geokémiai vizsgálata - petrogenetikai következtetések. Peltz, S., Vajdea, E., Balogh, K., Pécskay, Z., 1987. Contributions to the geochronological Földtani Közlöny 137 (1), 103–128. study of the volcanic processes int he Calimani and Hargitha Mountains (East Wenzel, F., Lorenz, F.P., Sperner, B., Oncescu, M.C., 1999. Seismotectonics of the Romanian Carpathians, Romania). Dari de Seama ale Sedintelor Institutul de Geologie si Vrancea Area. In: Wenzel, F., Lungu, D., Novak, O. (Eds.), Vrancea Earthquakes: Tec- Geofizica 72-73, 323–338. tonics, Hazard and Risk Mitigation: Contributions From the First International Work- Popa, M., Radulian, M., Szakács, A., Seghedi, I., Zaharia, B., 2012. New seismic and tomog- shop on Vrancea Earthquakes, Bucharest, Romania, November 1–4, 1997. Springer raphy data in the southern part of the Harghita Mountains (Romania, Southeastern Netherlands, Dordrecht:pp. 15–25 https://doi.org/10.1007/978-94-011-4748-4_2. Carpathians): connection with recent volcanic activity. Pure Appl. Geophys. 169 (9):1557–1573. https://doi.org/10.1007/s00024-011-0428-6.